Plasma membrane robertson_jd[1]
Dilippandya
1,385 views
16 slides
Apr 24, 2016
Slide 1 of 16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
About This Presentation
plasma membrane
Slide Content
MembraneStructure
J .DAVID ROBERTSON
Thischaptersurveysselectedhighlightsoftheevolutionof
modemideasaboutthemolecularorganization ofbiological
membranes.Thesurveyisinnosensecomplete,andreferences
1-14maybe consultedformoredetailsonthetopicscovered
here.Manyimportanttopicsleftalmostorcompletelyun-
touchedincludemembranetransport(15-21),blacklipidfilms
(22-31),andmanyaspectsofmembranebiochemistry(32-35).
HistoricalBackgroundBeforethe
ElectronMicroscope
EARLYIDEA S:
Theexistenceofsomekindofmembrane
structurethatboundscellswasimplicitlyrecognizedassoonas
thecellconceptwasdefinedbySchleidenandSchwannin
1839(36)
.Bowmanin1840(37)was oneofthefirsttodepict
suchastructureasananatomicalentityinhisdrawingsofthe
sarcolemma
.Theearliestintimationsthatthemembranecon-
tainedlipidcamefromtheworkofOvertonin1895(38,39).
Theessentialpointwasthe discoverythatlipid-solublemole-
culespenetratedintocellsmoreeasily.J.Bernsteindeveloped
thehypothesis,definitivelypresentedin1902(40)butintimated
asearlyas1868(41),thatlivingcellsconsistedofanelectrolyte
interiorsurroundedbyathinmembranerelativelyimperme-
abletoions
.Healsopostulatedthattherewasanelectrical
potentialdifferenceacross themembraneatrest,andthat
duringactivitythe ion permeabilitybarrierwasreducedtoa
relativelylowvalue.Theproofoftheessentialcorrectnessof
Bernstein'smainpointcamein1910--1913withtheexperiments
ofHober(42-44),whomeasuredtheelectricalresistancetoan
alternatingcurrentofamassof redbloodcellscentrifugedin
sucrose.Hefoundthatat1kilocycle/stheresistancewashigh
(-1,200SZcm) butthatitbecamemuchlower(-200Stcm)at
10megacycles/s
.Thelatteristheresistanceofa 0.4%NaCl
solution.Afterhemolysisandtreatmentwithsaponin,thesame
lowresistancewasfoundatbothlowandhighfrequenices.
Frickein1923(45,46)measuredthe capacitanceofthered
bloodcellmembranetobe0.81tAF/cm2.Hesupposedthe
membranetobe anoilfilmwithadielectricconstantof3and
socalculatedthethicknesstobe 33A.Thiswasthefirst
indicationthatamembranemightbe ofmoleculardimensions
.
Manymembranes,thoughnotall,weresubsequentlyfoundto
havea capacitanceof-1uF/cm
2,afirstintimationof the
existenceofsomekindofunitarystructure.
I.DAVID ROBERTSON DepartmentofAnatomy,DukeUniversity
MedicalCenter,Durham,NorthCarolina
THEJOURNALOFCELLBIOLOGY"VOLUME91NO.3 PT.2DECEMBER1981189s-204s
©TheRockefellerUniversity Press-0021-9525/81/12/189s/16$1.00
M ONOM OL ECULAIRFILMS:
Studiesonmonomolecular
filmswereoffundamentalimportance.LordRayleighin1890
(47)measuredthethicknessofafilmofoliveoiltobe1.63x
10 -7cm.Devaux(48)didmuchpioneeringwork onoilfilms
onwateraswellasproteinmonolayersatbothair-waterand
oil-waterinterfaces
.Langmuir(49) in1917(cf.Harkins[50-
51])showedthatsomelipidmoleculeswereamphiphilic(52),
inhavinga polarheadandanonpolarcarbonchain
.When
spreadatanair-waterinterface,theyformedamonolayeron
thesurfaceandaffectedthesurface tension.Whenasurface
barrierwasmovedsoas toreducetheareawhilemeasuring
force,acharacteristicforce/area curvewasobtained
.Amini-
malareawasreachedatwhichtheforcewasmaximalandthen
thefilmcollapsed,asevidencedbyabreakinthecurve.
Langmuirinterpretedthesefindingscorrectlyasshowingthat
thelipidmoleculeswereamphiphilic.Atalowsurface pressure
theywererandomlyarrangedinthewatersurface,butupon
beingpushedclosetogetherbythemovingbarrier,theyformed
astructurelikeapicketfence withtheirpolarheadsinthe
waterandtheirnonpolarcarbonchainspointedintotheair.
Hecalculatedthearea/moleculeofavarietyoffattyacidsand
othersubstances.
THE
LIPID
BILAYER:
In 1925GoiterandGrendel(53)
extractedthelipidfromaknownnumberofredbloodcells,
calculatedthetotalcellarea,andfoundthemeasuredminimal
area ofthetotallipidscompressedonamonolayertroughto
betwicethisvalue
.Thisledtothebilayerconcept.The
extractionproceduredidnotextractallof thelipids,butthis
wascompensatedbyunderestimationofthecellarea
.This
workprovidedthefirstsuggestionthatalipidbilayermightbe
afundamentalfeatureofbiologicalmembranes,butnoeffort
wasmadetogeneralize.
Schmittetal.in1937-1938(54,55)studiederythrocyte ghost
membranesinpolarizedlightandconcludedthatthey con-
tainedlipidmoleculesorientedperpendiculartotheplaneof
themembraneaswouldbeexpectedifalipidbilayerwere
present.
BIOPHYSICAL PROPERTIES OFMEMBRANES
Fluidity
Duringthe1930s,Chambersand Kopac(56,57)showed
thatanoildropletappliedtothesurfaceofadenudedmarine
eggquicklypassedthroughthesurfaceandappearedasa
dropletonthecytoplasmicside.Theyalsonotedthat,iftwo
suchdropletswereappliedondifferentareasandtheseawater
189S
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
J .DAVID ROBERTSON
Thischaptersurveysselectedhighlightsoftheevolutionof
modemideasaboutthemolecularorganization ofbiological
membranes.Thesurveyisinnosensecomplete,andreferences
1-14maybe consultedformoredetailsonthetopicscovered
here.Manyimportanttopicsleftalmostorcompletelyun-
touchedincludemembranetransport(15-21),blacklipidfilms
(22-31),andmanyaspectsofmembranebiochemistry(32-35).
HistoricalBackgroundBeforethe
ElectronMicroscope
EARLYIDEA S:
Theexistenceofsomekindofmembrane
structurethatboundscellswasimplicitlyrecognizedassoonas
thecellconceptwasdefinedbySchleidenandSchwannin
1839(36)
.Bowmanin1840(37)was oneofthefirsttodepict
suchastructureasananatomicalentityinhisdrawingsofthe
sarcolemma
.Theearliestintimationsthatthemembranecon-
tainedlipidcamefromtheworkofOvertonin1895(38,39).
Theessentialpointwasthe discoverythatlipid-solublemole-
culespenetratedintocellsmoreeasily.J.Bernsteindeveloped
thehypothesis,definitivelypresentedin1902(40)butintimated
asearlyas1868(41),thatlivingcellsconsistedofanelectrolyte
interiorsurroundedbyathinmembranerelativelyimperme-
abletoions
.Healsopostulatedthattherewasanelectrical
potentialdifferenceacross themembraneatrest,andthat
duringactivitythe ion permeabilitybarrierwasreducedtoa
relativelylowvalue.Theproofoftheessentialcorrectnessof
Bernstein'smainpointcamein1910--1913withtheexperiments
ofHober(42-44),whomeasuredtheelectricalresistancetoan
alternatingcurrentofamassof redbloodcellscentrifugedin
sucrose.Hefoundthatat1kilocycle/stheresistancewashigh
(-1,200SZcm) butthatitbecamemuchlower(-200Stcm)at
10megacycles/s
.Thelatteristheresistanceofa 0.4%NaCl
solution.Afterhemolysisandtreatmentwithsaponin,thesame
lowresistancewasfoundatbothlowandhighfrequenices.
Frickein1923(45,46)measuredthe capacitanceofthered
bloodcellmembranetobe0.81tAF/cm2.Hesupposedthe
membranetobe anoilfilmwithadielectricconstantof3and
socalculatedthethicknesstobe 33A.Thiswasthefirst
indicationthatamembranemightbe ofmoleculardimensions
.
Manymembranes,thoughnotall,weresubsequentlyfoundto
havea capacitanceof-1uF/cm
2,afirstintimationof the
existenceofsomekindofunitarystructure.
I.DAVID ROBERTSON DepartmentofAnatomy,DukeUniversity
MedicalCenter,Durham,NorthCarolina
THEJOURNALOFCELLBIOLOGY"VOLUME91NO.3 PT.2DECEMBER1981189s-204s
©TheRockefellerUniversity Press-0021-9525/81/12/189s/16$1.00
M ONOM OL ECULAIRFILMS:
Studiesonmonomolecular
filmswereoffundamentalimportance.LordRayleighin1890
(47)measuredthethicknessofafilmofoliveoiltobe1.63x
10 -7cm.Devaux(48)didmuchpioneeringwork onoilfilms
onwateraswellasproteinmonolayersatbothair-waterand
oil-waterinterfaces
.Langmuir(49) in1917(cf.Harkins[50-
51])showedthatsomelipidmoleculeswereamphiphilic(52),
inhavinga polarheadandanonpolarcarbonchain
.When
spreadatanair-waterinterface,theyformedamonolayeron
thesurfaceandaffectedthesurface tension.Whenasurface
barrierwasmovedsoas toreducetheareawhilemeasuring
force,acharacteristicforce/area curvewasobtained
.Amini-
malareawasreachedatwhichtheforcewasmaximalandthen
thefilmcollapsed,asevidencedbyabreakinthecurve.
Langmuirinterpretedthesefindingscorrectlyasshowingthat
thelipidmoleculeswereamphiphilic.Atalowsurface pressure
theywererandomlyarrangedinthewatersurface,butupon
beingpushedclosetogetherbythemovingbarrier,theyformed
astructurelikeapicketfence withtheirpolarheadsinthe
waterandtheirnonpolarcarbonchainspointedintotheair.
Hecalculatedthearea/moleculeofavarietyoffattyacidsand
othersubstances.
THE
LIPID
BILAYER:
In 1925GoiterandGrendel(53)
extractedthelipidfromaknownnumberofredbloodcells,
calculatedthetotalcellarea,andfoundthemeasuredminimal
area ofthetotallipidscompressedonamonolayertroughto
betwicethisvalue
.Thisledtothebilayerconcept.The
extractionproceduredidnotextractallof thelipids,butthis
wascompensatedbyunderestimationofthecellarea
.This
workprovidedthefirstsuggestionthatalipidbilayermightbe
afundamentalfeatureofbiologicalmembranes,butnoeffort
wasmadetogeneralize.
Schmittetal.in1937-1938(54,55)studiederythrocyte ghost
membranesinpolarizedlightandconcludedthatthey con-
tainedlipidmoleculesorientedperpendiculartotheplaneof
themembraneaswouldbeexpectedifalipidbilayerwere
present.
BIOPHYSICAL PROPERTIES OFMEMBRANES
Fluidity
Duringthe1930s,Chambersand Kopac(56,57)showed
thatanoildropletappliedtothesurfaceofadenudedmarine
eggquicklypassedthroughthesurfaceandappearedasa
dropletonthecytoplasmicside.Theyalsonotedthat,iftwo
suchdropletswereappliedondifferentareasandtheseawater
189S
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
wasagitated,thedropletsmovedrelativeto oneanother,
indicatingthatthemembranewasfluid(58).
SurfaceProperties
MuddandMuddin 1931(59)did arevealingexperimenton
redbloodcells(cf.60).Theyexaminedmicroscopicallya
dropletofbloodonaglassslideincontact withanoildroplet
underacoverslip.Thewhitecellsremainedin thewaterphase,
butsomeredbloodcellsentered theoilphase,showingthat
theirsurfaceswererelativelyhydrophobic.Theseobservations
indicatedthattheexternalsurfaceoftheerythrocytemembrane
wascoveredbymaterial,probablyprotein,whichcouldbecome
predominantlyhydrophobic.Thismeansthathydrophobic
bondingof aproteintoamembranesurfacedoesnotneces-
sarilyrequirebondingtolipid.
In 1932Cole(61)measuredthesurfacetensionofstarfish
eggmembranes.Hedeterminedtheforcerequiredtocompress
aneggbetweentwoglasscoverslipsandcalculatedsurface
tension
valuesof-0.1dyn/cm.Harvey andShapiro(62)found
similarvaluesbymeasuringthesurfacetension ofoildroplets
withincellswiththecentrifugemicroscope.Theselowvalues
seemedstrangebecausepeoplewerethinking ofcellmem-
branesasthin,oilyfilmsandthesurface tensionvaluesofoils
weremuchhigher.Danielliand Harveyin1934(63)studied
oildropletsfrommackereleggsandfoundthatafterextensive
washingtheygavesurface tensionvalues ofabout9dyn/cm.
Whenacytoplasmicextractwasaddedtotheoil,thesurface
tensionwasloweredandtheyidentifiedthe agentresponsible
asprotein.
TheDevauxEffect
In 1938LangmuirandWaugh(64)gavethename"Devaux
effect"toaphenomenonrelatedtotheabove.Devaux(65)
simplyshookasolutionofalbumeninwaterwithbenzeneand
notedthatatacertainalbumenconcentrationthebenzene
formeddropletsandthealbumenspreadat the'oil-water
interfaceintoamonolayer.Thesurfacetensionatthe resulting
interface
wasobviouslyverylowbecausethedropletssponta-
neouslyassumedpeculiarshapes.Similarly,Danielli(66)found
that proteinsspreadatanoil-waterinterfaceshowedaninitial
markedfallinsurfacetension,whichrosewith timetoafinal
valuelessthanthatoftheoil-waterinterfacealone.
Kopac(67-68) reportedthata droplet ofoilinjectedintoa
protein solutionsoonbecamecrenated,indicatingthedevel-
opmentoflowsurface tensionbecauseof theDevauxeffect.
Dropletsmicroinjectedintothecytoplasmofmarineeggcells
remainedsmooth andspherical.However,ifthecellwas
pricked with aneedletocausecytolysis,theoildropletimme-
diatelybecamecrenated.Later Trurnit(69)madetherelevant
pointtheproteinsinsolutiongenerallyadsorbandspreadat
theair-waterinterfaceaswellasanyotherhigh-tensionsurface.
Thefactthatnosuchinterfacialspreading occurredin the
intacteggindicatedthatthe eggcytoplasmicmatrixdid not
contain proteinmoleculesinsimplesolution.
Alltheseexperimentswereimportantintheearlyevolution
ofthinkingaboutmembranestructure.Interestingly,theim-
portantpointwasmissedthatsomenaturalphospholipids,e.g.,
phosphatidylcholine(PC),inmonolayersgivequitelowsur-
face-tension valuesof<5dyn/cm(70).Synthetic dipalmitoyl
lecithin atanair-waterinterfacegivesasurface-tensionvalue
thatishardlymeasurable(0-2dyn/cm)(70,71).Phospholipids
arethedominantlipidsofbiologicalmembranesand oneof
1905
THEJOURNALOFCELLBIOLOGY "VOLUME91,1981
theirfunctionsmaybeto conferlowsurface-tension properties
onmembranelipids.Thiscouldbeimportantinpreventingthe
denaturationofmembraneproteins.
MembraneCapacitance
In1950 ColeandCurtis(72)tabulatedthe knownvaluesof
membranecapacitance,rangingfrom0.81
AF/cm'
forthe
humanerythrocytemembranethrough9.0,uF/cm2forcow
erythrocytesto0.0121uF/cm2forfrogperonealnerve.Colein
1935
(73)madethesurprisingobservationthatseaurchineggs,
normallyhavinga capacitance of1gF/cm2,displayedlower
valueswhenswollen.Heexpectedthereversebecauseof the
thinning
ofthemembrane.Instead,thelowervalues suggested
thickening
.Hisfindingswereconfirmedbylida(74,75)who
showedthephenomenontobereversible.Highcapacitance
valueswerealsoobtainedforskeletalmusclefibersby Katz
(76).
Thereasonforsomeofthevariationsinmembranecapaci-
tancewaselucidatedbyLordRothschildin1957(77).He
showedthattherewasanerrorinthecalculationofthe areaof
eggmembranes.Byelectronmicroscopyhefoundthatthe
surfacemembranewasthrownintominutefoldsnotapparent
bylightmicroscopy,thusincreasinggreatlythe actualsurface
area.Similarly,withourunderstandingoftheTsystemin
skeletal
musclefibers(78-81),itbecameknowninthelate
1950sthattheactualmeasuredareas ofmusclefibersusedin
calculatingmembranecapacitancewerewrongbecausetheT
systemwasnottakenintoaccount.
THEDANIELLI-DAVSONMODEL:In1935,Danielli
andDavson(82)presentedamodelofcellmembranestructure
whichtheylatergeneralized(58) intothe"pauci-molecular"
theorythatstatedthatallbiologicalmembraneshada"lipoid"
coreborderedbymonolayersoflipidwiththelipidpolarheads
pointedoutwardandcoveredbyproteinmonolayers(Fig.1).
In 1943 (Fig.16binreference58) they presentedadetailed
moleculardiagram showinghydrophobicaminoacidside-
chainspenetratingbetweenlipidheadgroupsandlyingbetween
thecarbonchainsoflipidmoleculesinarelationshipthat
wouldresult inhydrophobicbonding.
Thepauci-moleculartheoryprovidedthemajormembrane
paradigmintothe 1950s.Theoriginaltheorydidnotspecify
thebilayerasa generalstructure,althoughDanielliclearly
favoreditand,inoneofhislaterdiagrams(83),he drew
transmembrane
polypeptidechainsarrangedwiththeirpolar
groupsapposedsoastomakea polartransmembranechannel.
Inapaperpublishedin1935onthethicknessofthered
bloodcellmembrane,Danielli (84)discussedFricke'schoice
ofa valueof3for thedielectricconstant.Henotedthatthe
valuedependsontheorientationofthedipolesinthefilmand
indicatedthathighervalues,upto6.7,could occur.Thiswould
requirealargerthicknessvalueforthemembranefora given
measuredcapacitance.Perhapsthisiswhythemodelwas
thickerthanonebilayer.In1936,Danielli(85)drewanumber
of modelsvaryingfromasinglelipidmonolayertothethicker
originalmodel.Thecautiouswayinwhichthisproblemwas
treatedillustratestheuncertaintyaboutmembranethickness
oftheperiod.
In1949,WaughandSchmitt(86)measuredthethicknessof
erythrocytemembraneswiththe"analyticalleptoscope."Their
results,thoughcompatiblewithasingle bilayer,werenot
unambiguousandwererestrictedtoerythrocytemembranes.
Thus,itseemsfair tosaythatthesinglebilayermodel,which
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
indicatingthatthemembranewasfluid(58).
SurfaceProperties
MuddandMuddin 1931(59)did arevealingexperimenton
redbloodcells(cf.60).Theyexaminedmicroscopicallya
dropletofbloodonaglassslideincontact withanoildroplet
underacoverslip.Thewhitecellsremainedin thewaterphase,
butsomeredbloodcellsentered theoilphase,showingthat
theirsurfaceswererelativelyhydrophobic.Theseobservations
indicatedthattheexternalsurfaceoftheerythrocytemembrane
wascoveredbymaterial,probablyprotein,whichcouldbecome
predominantlyhydrophobic.Thismeansthathydrophobic
bondingof aproteintoamembranesurfacedoesnotneces-
sarilyrequirebondingtolipid.
In 1932Cole(61)measuredthesurfacetensionofstarfish
eggmembranes.Hedeterminedtheforcerequiredtocompress
aneggbetweentwoglasscoverslipsandcalculatedsurface
tension
valuesof-0.1dyn/cm.Harvey andShapiro(62)found
similarvaluesbymeasuringthesurfacetension ofoildroplets
withincellswiththecentrifugemicroscope.Theselowvalues
seemedstrangebecausepeoplewerethinking ofcellmem-
branesasthin,oilyfilmsandthesurface tensionvaluesofoils
weremuchhigher.Danielliand Harveyin1934(63)studied
oildropletsfrommackereleggsandfoundthatafterextensive
washingtheygavesurface tensionvalues ofabout9dyn/cm.
Whenacytoplasmicextractwasaddedtotheoil,thesurface
tensionwasloweredandtheyidentifiedthe agentresponsible
asprotein.
TheDevauxEffect
In 1938LangmuirandWaugh(64)gavethename"Devaux
effect"toaphenomenonrelatedtotheabove.Devaux(65)
simplyshookasolutionofalbumeninwaterwithbenzeneand
notedthatatacertainalbumenconcentrationthebenzene
formeddropletsandthealbumenspreadat the'oil-water
interfaceintoamonolayer.Thesurfacetensionatthe resulting
interface
wasobviouslyverylowbecausethedropletssponta-
neouslyassumedpeculiarshapes.Similarly,Danielli(66)found
that proteinsspreadatanoil-waterinterfaceshowedaninitial
markedfallinsurfacetension,whichrosewith timetoafinal
valuelessthanthatoftheoil-waterinterfacealone.
Kopac(67-68) reportedthata droplet ofoilinjectedintoa
protein solutionsoonbecamecrenated,indicatingthedevel-
opmentoflowsurface tensionbecauseof theDevauxeffect.
Dropletsmicroinjectedintothecytoplasmofmarineeggcells
remainedsmooth andspherical.However,ifthecellwas
pricked with aneedletocausecytolysis,theoildropletimme-
diatelybecamecrenated.Later Trurnit(69)madetherelevant
pointtheproteinsinsolutiongenerallyadsorbandspreadat
theair-waterinterfaceaswellasanyotherhigh-tensionsurface.
Thefactthatnosuchinterfacialspreading occurredin the
intacteggindicatedthatthe eggcytoplasmicmatrixdid not
contain proteinmoleculesinsimplesolution.
Alltheseexperimentswereimportantintheearlyevolution
ofthinkingaboutmembranestructure.Interestingly,theim-
portantpointwasmissedthatsomenaturalphospholipids,e.g.,
phosphatidylcholine(PC),inmonolayersgivequitelowsur-
face-tension valuesof<5dyn/cm(70).Synthetic dipalmitoyl
lecithin atanair-waterinterfacegivesasurface-tensionvalue
thatishardlymeasurable(0-2dyn/cm)(70,71).Phospholipids
arethedominantlipidsofbiologicalmembranesand oneof
1905
THEJOURNALOFCELLBIOLOGY "VOLUME91,1981
theirfunctionsmaybeto conferlowsurface-tension properties
onmembranelipids.Thiscouldbeimportantinpreventingthe
denaturationofmembraneproteins.
MembraneCapacitance
In1950 ColeandCurtis(72)tabulatedthe knownvaluesof
membranecapacitance,rangingfrom0.81
AF/cm'
forthe
humanerythrocytemembranethrough9.0,uF/cm2forcow
erythrocytesto0.0121uF/cm2forfrogperonealnerve.Colein
1935
(73)madethesurprisingobservationthatseaurchineggs,
normallyhavinga capacitance of1gF/cm2,displayedlower
valueswhenswollen.Heexpectedthereversebecauseof the
thinning
ofthemembrane.Instead,thelowervalues suggested
thickening
.Hisfindingswereconfirmedbylida(74,75)who
showedthephenomenontobereversible.Highcapacitance
valueswerealsoobtainedforskeletalmusclefibersby Katz
(76).
Thereasonforsomeofthevariationsinmembranecapaci-
tancewaselucidatedbyLordRothschildin1957(77).He
showedthattherewasanerrorinthecalculationofthe areaof
eggmembranes.Byelectronmicroscopyhefoundthatthe
surfacemembranewasthrownintominutefoldsnotapparent
bylightmicroscopy,thusincreasinggreatlythe actualsurface
area.Similarly,withourunderstandingoftheTsystemin
skeletal
musclefibers(78-81),itbecameknowninthelate
1950sthattheactualmeasuredareas ofmusclefibersusedin
calculatingmembranecapacitancewerewrongbecausetheT
systemwasnottakenintoaccount.
THEDANIELLI-DAVSONMODEL:In1935,Danielli
andDavson(82)presentedamodelofcellmembranestructure
whichtheylatergeneralized(58) intothe"pauci-molecular"
theorythatstatedthatallbiologicalmembraneshada"lipoid"
coreborderedbymonolayersoflipidwiththelipidpolarheads
pointedoutwardandcoveredbyproteinmonolayers(Fig.1).
In 1943 (Fig.16binreference58) they presentedadetailed
moleculardiagram showinghydrophobicaminoacidside-
chainspenetratingbetweenlipidheadgroupsandlyingbetween
thecarbonchainsoflipidmoleculesinarelationshipthat
wouldresult inhydrophobicbonding.
Thepauci-moleculartheoryprovidedthemajormembrane
paradigmintothe 1950s.Theoriginaltheorydidnotspecify
thebilayerasa generalstructure,althoughDanielliclearly
favoreditand,inoneofhislaterdiagrams(83),he drew
transmembrane
polypeptidechainsarrangedwiththeirpolar
groupsapposedsoastomakea polartransmembranechannel.
Inapaperpublishedin1935onthethicknessofthered
bloodcellmembrane,Danielli (84)discussedFricke'schoice
ofa valueof3for thedielectricconstant.Henotedthatthe
valuedependsontheorientationofthedipolesinthefilmand
indicatedthathighervalues,upto6.7,could occur.Thiswould
requirealargerthicknessvalueforthemembranefora given
measuredcapacitance.Perhapsthisiswhythemodelwas
thickerthanonebilayer.In1936,Danielli(85)drewanumber
of modelsvaryingfromasinglelipidmonolayertothethicker
originalmodel.Thecautiouswayinwhichthisproblemwas
treatedillustratestheuncertaintyaboutmembranethickness
oftheperiod.
In1949,WaughandSchmitt(86)measuredthethicknessof
erythrocytemembraneswiththe"analyticalleptoscope."Their
results,thoughcompatiblewithasingle bilayer,werenot
unambiguousandwererestrictedtoerythrocytemembranes.
Thus,itseemsfair tosaythatthesinglebilayermodel,which
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
FIGURE1Schematic diagramofthemolecularconditionsatthe
cellsurfacepublishedbyDanielliand Davson(82) in1935.
Danielliclearlyfavored,wasbasedongoodevidencefor the
erythrocytemembrane,whereasnohardevidenceforagener-
alizationofthebilayerexisted.
TheEarlyElectronMicroscopePeriod
THEUNITMEMBRANEMODEL:
Inthe1950stheelec-
tronmicroscopemadeitpossibletolookprofitablyatsectioned
cellsatresolutionsbetterthan50A.Theintroductionofpotas-
siumpermanganateasafixingagentbyLuft(87)andepoxy
resinsasembeddingmaterialsbyGlauertet al.(88)ledto
visualizationofthecellmembraneasatriple-layeredstructure
-75Athickconsisting oftwodensestrata,eachabout25A
thick,borderingalightcentralzone ofabout equalthickness.
Thistriple-layeredpatternwasobservedinnervefibers(89-
92)andinothertissues(93-97).
Figure2isanelectronmicrographofahumanerythrocyte
membranefixedwithglutaraldehyde,embeddedinpolyglutar-
aldehydebytheglutaraldehyde-carbohydrazide(GACH)
method(98),andstainedwithuranylandleadsalts.The
overallthicknessis-100A.Thebilayercoreistheclearstratum
-r40Athickbetweenthetwodensesurfacestrata.Thecoreof
the
bilayerdoesnottakeupheavy-metalstrainsbecauseit is
hydrophobic.Itconsistsmainlyofthehydrocarbonchainsof
lipidmoleculesandthehydrophobicpolypeptidechains of
integralmembraneproteins.Thesurfacestrataarehydrophilic
andtakeupthestainsavidly.Thehydrophilicstructuresare
the polarheadsoflipidmolecules,proteinmolecules,and,in
the outsidesurface,carbohydrateresidues.Asymmetryisnot
revealed
bythismethod.
Asimilar,butthinner,triple-layeredpatternwasalsoob-
served
inmodelsystemsconsistingofsmectic fluidcrystalsof
phospholipids(94-96).Whenthesewerehydrated, theindivid-
ualbilayersappearedaspairsofdensestrataseparatedbya
lightcentralzone,in thiswaylookingverymuchlike cell
membranesbutdefinitelythinner.Thisfindingmeantthat
heavy-metalatomsaccumulatedinthe polarheadregionsof
thelipidmoleculesalthoughtheprimaryreactionofOS04was
withthedoublebondsofthelipidcarbonchains(99).Single
bilayersdidnotappearassingle,densestratabut alwaysas a
pairofdensestratamakingatriple-layeredstructure.Thiswas
rationalized(5,94-97)byassumingthattheprimaryreaction
productof the Criegeereaction,OSO3,becauseit ismorepolar
thanOS04,must bedrivenoutofthehydrophobicinteriorof
themembraneandbecomeadsorbedinthepolarregions,
increasingtherelativedensitytherebyaddingtodensitycaused
bydirectreactionofheadgroupsconstituents.Stoeckenius
(100)laterperformedsomeexperimentsbasedontheworkof
LuzattiandHusson(101)onmodellipids,andconfirmedthis
interpretation.
In1954,Geren(102)postulatedthatthenervemyelinsheath
mightconsistof aspirallywrapped mesaxon;in 1955(103),
bothouterandinnermesaxonswereobservedtoconnecta
fullydeveloped myelinsheathwiththeinnerandoutersurfaces
oftheSchwanncell,provinghertheory.Theapplicationof
permanganatefixationandepoxyembeddingtodeveloping
mousesciaticnervefibersshowedthemesaxonclearlyastwo
triple-layeredmembranesunitedalongtheirexternalsurfaces,
andalsoshowedthatcompactmyelinresultedfromtheclose
appositionofthecytoplasmicsurfacesofthemembranesofthe
mesaxon.Atthetime, therewasatentativemolecularmodel
oftheradiallyrepeatingunit ofthemyelinsheath(104-106).
Inasmuchasthemesaxonwasobviouslythe repeatingunit,it
waspossibletoidentifythestratawithincompactmyelinin
molecular terms(92-95).Thisanalysisindicatedthatthe
Schwanncellmembranewasalipidbilayercoveredby mono-
layersofnonlipidsoneitherside,asinFig.3.Thevalidityof
thisanalysiswasconfirmedlaterbyX-raydiffractionstudies
whenthephase problemwassolvedataresolutionof-30Aby
Moody(107)andbyCasparandKirschner(108)athigher
resolution.
Itwasalso possibletodeducefromthestructureofthe
myelinsheath,aswellasfromitsstainingcharacteristicsob-
served
byelectronmicroscopy,thatthe outersurfaceofthe
membranewaschemicallydifferentfromtheinnersurface.
Twounitmembraneswereincludedinoneradialrepeating
unit,whichshowedthattherehadtobeadifferencebetween
theinsideandoutsidesurfacesofthemembrane,Finean's
"differencefactor"(105).Therewasagoodreasonforgener-
alizationofthe idea ofmembraneasymmetry.Theexternal
FIGURE2
Electronmicrographofahumanerythrocytemembrane
fixedwithglutaraldehydeembeddedinpolyglutaraldehydebythe
GACHmethod(98)andstainedwithuranylandleadsalts.The
membraneconsistsoftwodensestrataseparatedby alightcentral
zone.Theoverallthicknessisabout130-140Ain thispreparation.
ThisishigherthanisseenafterOS04fixationandthehighvalueis
probablyduetosomedisplacementsduetosectioningasiscommon
inGACHembeddedspecimensnotpostfixedwithOs04andRu04.
x100,000
ROBERTSON
MembraneStructure
191S
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
cellsurfacepublishedbyDanielliand Davson(82) in1935.
Danielliclearlyfavored,wasbasedongoodevidencefor the
erythrocytemembrane,whereasnohardevidenceforagener-
alizationofthebilayerexisted.
TheEarlyElectronMicroscopePeriod
THEUNITMEMBRANEMODEL:
Inthe1950stheelec-
tronmicroscopemadeitpossibletolookprofitablyatsectioned
cellsatresolutionsbetterthan50A.Theintroductionofpotas-
siumpermanganateasafixingagentbyLuft(87)andepoxy
resinsasembeddingmaterialsbyGlauertet al.(88)ledto
visualizationofthecellmembraneasatriple-layeredstructure
-75Athickconsisting oftwodensestrata,eachabout25A
thick,borderingalightcentralzone ofabout equalthickness.
Thistriple-layeredpatternwasobservedinnervefibers(89-
92)andinothertissues(93-97).
Figure2isanelectronmicrographofahumanerythrocyte
membranefixedwithglutaraldehyde,embeddedinpolyglutar-
aldehydebytheglutaraldehyde-carbohydrazide(GACH)
method(98),andstainedwithuranylandleadsalts.The
overallthicknessis-100A.Thebilayercoreistheclearstratum
-r40Athickbetweenthetwodensesurfacestrata.Thecoreof
the
bilayerdoesnottakeupheavy-metalstrainsbecauseit is
hydrophobic.Itconsistsmainlyofthehydrocarbonchainsof
lipidmoleculesandthehydrophobicpolypeptidechains of
integralmembraneproteins.Thesurfacestrataarehydrophilic
andtakeupthestainsavidly.Thehydrophilicstructuresare
the polarheadsoflipidmolecules,proteinmolecules,and,in
the outsidesurface,carbohydrateresidues.Asymmetryisnot
revealed
bythismethod.
Asimilar,butthinner,triple-layeredpatternwasalsoob-
served
inmodelsystemsconsistingofsmectic fluidcrystalsof
phospholipids(94-96).Whenthesewerehydrated, theindivid-
ualbilayersappearedaspairsofdensestrataseparatedbya
lightcentralzone,in thiswaylookingverymuchlike cell
membranesbutdefinitelythinner.Thisfindingmeantthat
heavy-metalatomsaccumulatedinthe polarheadregionsof
thelipidmoleculesalthoughtheprimaryreactionofOS04was
withthedoublebondsofthelipidcarbonchains(99).Single
bilayersdidnotappearassingle,densestratabut alwaysas a
pairofdensestratamakingatriple-layeredstructure.Thiswas
rationalized(5,94-97)byassumingthattheprimaryreaction
productof the Criegeereaction,OSO3,becauseit ismorepolar
thanOS04,must bedrivenoutofthehydrophobicinteriorof
themembraneandbecomeadsorbedinthepolarregions,
increasingtherelativedensitytherebyaddingtodensitycaused
bydirectreactionofheadgroupsconstituents.Stoeckenius
(100)laterperformedsomeexperimentsbasedontheworkof
LuzattiandHusson(101)onmodellipids,andconfirmedthis
interpretation.
In1954,Geren(102)postulatedthatthenervemyelinsheath
mightconsistof aspirallywrapped mesaxon;in 1955(103),
bothouterandinnermesaxonswereobservedtoconnecta
fullydeveloped myelinsheathwiththeinnerandoutersurfaces
oftheSchwanncell,provinghertheory.Theapplicationof
permanganatefixationandepoxyembeddingtodeveloping
mousesciaticnervefibersshowedthemesaxonclearlyastwo
triple-layeredmembranesunitedalongtheirexternalsurfaces,
andalsoshowedthatcompactmyelinresultedfromtheclose
appositionofthecytoplasmicsurfacesofthemembranesofthe
mesaxon.Atthetime, therewasatentativemolecularmodel
oftheradiallyrepeatingunit ofthemyelinsheath(104-106).
Inasmuchasthemesaxonwasobviouslythe repeatingunit,it
waspossibletoidentifythestratawithincompactmyelinin
molecular terms(92-95).Thisanalysisindicatedthatthe
Schwanncellmembranewasalipidbilayercoveredby mono-
layersofnonlipidsoneitherside,asinFig.3.Thevalidityof
thisanalysiswasconfirmedlaterbyX-raydiffractionstudies
whenthephase problemwassolvedataresolutionof-30Aby
Moody(107)andbyCasparandKirschner(108)athigher
resolution.
Itwasalso possibletodeducefromthestructureofthe
myelinsheath,aswellasfromitsstainingcharacteristicsob-
served
byelectronmicroscopy,thatthe outersurfaceofthe
membranewaschemicallydifferentfromtheinnersurface.
Twounitmembraneswereincludedinoneradialrepeating
unit,whichshowedthattherehadtobeadifferencebetween
theinsideandoutsidesurfacesofthemembrane,Finean's
"differencefactor"(105).Therewasagoodreasonforgener-
alizationofthe idea ofmembraneasymmetry.Theexternal
FIGURE2
Electronmicrographofahumanerythrocytemembrane
fixedwithglutaraldehydeembeddedinpolyglutaraldehydebythe
GACHmethod(98)andstainedwithuranylandleadsalts.The
membraneconsistsoftwodensestrataseparatedby alightcentral
zone.Theoverallthicknessisabout130-140Ain thispreparation.
ThisishigherthanisseenafterOS04fixationandthehighvalueis
probablyduetosomedisplacementsduetosectioningasiscommon
inGACHembeddedspecimensnotpostfixedwithOs04andRu04.
x100,000
ROBERTSON
MembraneStructure
191S
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
FIGURE3Originalunitmembranemodel.Thelipidbilayerisin-
dicatedveryschematicallybythe barandcirclefigures.The non-
lipidmonolayersatthepolarsurfacesareindicatedbythe zigzag
lines.Thechemicalasymmetryproducedbythepresenceofcar-
bohydrateintheexternalsurfaceisindicatedby thepartialfillingin
ofthe zigzagrepresentingtheexternalmonolayer.
stratumof theSchwanncellmembraneinOSO4-fixedmyelin
oftenappearedtobefragmented.Atthefreesurfaceofthe
Schwanncell,as well asothercells,OsO4aloneusuallypre-
servedonlythecytoplasmicdensestratum,whereasKMn04
preservedbothstrata.Revelet al.(109)reportedthatglycogen
granulesinlivercellswerenotwellfixedwithOS04butwere
wellpreservedwithKMn04,fittingtheviewthattheouter
surfaceofmembranescontainedcarbohydrate(93-97).
Asurveyconductedofmanydifferenttissues,inseveral
differentanimalsindifferentphylaandbacteria,showedthat
thetriple-layeredpatterncouldbedemonstratedwithKMn04
inallcellularmembraneswhetheratthesurfaceor inmem-
branousorganelles(93).Itwasconcludedthatallbiological
membranesconsistedofthesamekindoffundamentalstruc-
turalpattern, i.e.,alipidbilayerarrangedwith the polarheads
ofthelipidmoleculespointingoutwardandcoveredbymono-
layersofnonlipid with apreponderanceofcarbohydrateinthe
externalsurface,as inFig.3.In thatthisstructurewasthe
repeatingunitofmyelinandofallmembranousstructuresof
cells,itwascalleda"unit"membrane.Theunit-membrane
theory(89,93-97,110)introducedanewparadigmthatwas
usefulforabout15years.Themodelbuiltontheearlier
Danielli-Davsonmodelbyadding twonewconcepts:itproved
theuniversalityofthesinglebilayerandintroducedforthe
firsttime the ideaofchemicalasymmetry,neitherofwhich
werefeaturesoftheearliermodel.Tobesure,Danielliclearly
believedthebilayertobethedominantstructureandhe
deservescreditforthis.Heevenguessedtheexistenceof
transmembraneproteins.
In1966,theunit-membrane paradigmcameunderattack
(111)
becauseitwasbelievedthatthestructureofcellmem-
branesmust bemorecomplicatedthanthetheoryseemedto
imply.Inpointing to thefactthatallmembraneshadthesame
kindofbasicstructural plan,theimpressionwasgiventhatall
membranesweremolecularlyidentical.Thiswas, ofcourse,a
complete misunderstanding(96).Theunit-membranemodel
wasincompleteinthatitdid not deal withmembranefluidity
norwiththe idea ofpenetratingproteins.Itwasdeficientin
that
itimpliedthatmembraneproteinswereunfoldedin the
samemannerasproteinsatair-waterinterfaces.However,its
majorfeatures-theuniversalityofthelipidbilayerandchem-
icalasymmetry-aregenerallyacceptedtoday.
THE SUBUNITM ODE LS:
Variousalternativemodelsin
1925
THEIOURNALOfCELLBIOLOGY"VOLUME91,1981
whichthebilayerwasalteredorinterruptedinavarietyof
wayswereproposedinthe1960sbySjbstrand(112,113),
LenardandSinger(114),Greenandhiscolleagues(115,116),
andBenson(117,118).Theessenceofthesemodelswasthat
thebilayerwasnot thedominantstructure,butthatlipid
moleculeswerearrangedin variouspatternsinthemembrane.
TheBensonmodelwasthemostextreme;themembrane
consistedofathinlayerofproteinwithlipidmoleculessimply
intercalatedinavarietyofways.Thepresent-day Sjbstrand
andBarajasmodels(119-121)aresomewhatsimilar.Theseall
fall
moreorlessintothe generalrubricof"subunit" models,
andtheearlieronesweredealtwithquitethoroughlyina
reviewin1969byStoeckeniusandEngleman(14),inwhich
theyconcludedthatthebilayermodelwastheonlyreasonable
one.Thebasicfactherewasthatitwasfoundimpossibleto
breakupanymembranestructureintosubunitsofuniform
compositionthatwouldreassembleintoafunctionalmem-
brane.Another problemwasthat thisconcept impliedthat
membranebiogenesisoccurredbyadditionsofaliquotsof
componentsasaggregatesofmanymoleculesthatservedas
buildingblocksforthewholestructure.Studiesofmembrane
biogenesisshouldthenhaveshownevidenceofparallelin-
creasesofatleastsomesetofrelatedcomponents.Nosuch
increaseswerefound.Forexample,Siekevitz,Palade,andco-
workers(122-128)showedthatindividualnascentmembrane
proteinsareinsertedintodevelopingmembranesystemsat
differenttimes,andthatproteinsinmembranesturn overat
differentrates,independentofoneanotherandindependent
oftheturnoverofmembranephospholipids.
TheModernPeriod
FREEZE-FRACTURE-ETCH(FIFE)ELECTRONMICROS-
CoP Y:Freeze-fracture-etchelectronmicroscopyhasbecome
importantinstructuralstudiesofmembranes.Steere
introducedtheFFEtechniquein1957(129),anditwasdevel-
opedfurtherin 1961byMoorandMuhlethaler(130),who
notedthatmembranesfractured transversely givethesame
triple-layeredunitmembraneappearancefoundinsections.
Theybelievedthatwhenthefractureplanefollowedthe plane
of
themembraneitranalongitssurface.Brantonandhis
associateslater(131-133)proposedthatfrozenbiologicalmem-
branestendtofracturecentrally,andnotedthatthefractured
replicatedsurfacesdisplayedparticlesabout50-100thindi-
ameter(Fig.4);daSilvaandBranton(134)provedthatthe
particleswerelocatedinside theerythrocytemembraneby
labelingtheexternalsurfaceswithferritinandidentifyingitin
theexternal etch(ES)facesinaplaneoutsidetheparticulate
fracture
faces.Theparticlesweremuchmore numerousonthe
protoplasmic(PF)thanontheexternalfracture(EF)facesof
erythrocytemembranes.Almostallfracturedmembranes
showedthisdistributionofparticles,buttheywerenotusually
observedinthenervemyelinsheathandinpurelipidmodel
systems (135-140).Pitsinthecomplementaryfracturefaces,
althoughsometimesseen,wereusuallyabsent.Generally,ret-
inalrodouter segmentmembranesalsofailed toshowdiscrete
particles(141-145).
THEFLUIDMOSAIC(FM)MODEL:
Singer(11,12)
andSingerandNicholsonin1972(146)proposedanew
membrane
paradigmwhichtheycalledthe"fluidmosaic"
model.Thisretainedthebilayerconcept,butintroducedanew
way
of lookingatthedistributionofprotein.Boththeouter
andinnersurfacesweredepicted aslargelynakedlipid.The
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
dicatedveryschematicallybythe barandcirclefigures.The non-
lipidmonolayersatthepolarsurfacesareindicatedbythe zigzag
lines.Thechemicalasymmetryproducedbythepresenceofcar-
bohydrateintheexternalsurfaceisindicatedby thepartialfillingin
ofthe zigzagrepresentingtheexternalmonolayer.
stratumof theSchwanncellmembraneinOSO4-fixedmyelin
oftenappearedtobefragmented.Atthefreesurfaceofthe
Schwanncell,as well asothercells,OsO4aloneusuallypre-
servedonlythecytoplasmicdensestratum,whereasKMn04
preservedbothstrata.Revelet al.(109)reportedthatglycogen
granulesinlivercellswerenotwellfixedwithOS04butwere
wellpreservedwithKMn04,fittingtheviewthattheouter
surfaceofmembranescontainedcarbohydrate(93-97).
Asurveyconductedofmanydifferenttissues,inseveral
differentanimalsindifferentphylaandbacteria,showedthat
thetriple-layeredpatterncouldbedemonstratedwithKMn04
inallcellularmembraneswhetheratthesurfaceor inmem-
branousorganelles(93).Itwasconcludedthatallbiological
membranesconsistedofthesamekindoffundamentalstruc-
turalpattern, i.e.,alipidbilayerarrangedwith the polarheads
ofthelipidmoleculespointingoutwardandcoveredbymono-
layersofnonlipid with apreponderanceofcarbohydrateinthe
externalsurface,as inFig.3.In thatthisstructurewasthe
repeatingunitofmyelinandofallmembranousstructuresof
cells,itwascalleda"unit"membrane.Theunit-membrane
theory(89,93-97,110)introducedanewparadigmthatwas
usefulforabout15years.Themodelbuiltontheearlier
Danielli-Davsonmodelbyadding twonewconcepts:itproved
theuniversalityofthesinglebilayerandintroducedforthe
firsttime the ideaofchemicalasymmetry,neitherofwhich
werefeaturesoftheearliermodel.Tobesure,Danielliclearly
believedthebilayertobethedominantstructureandhe
deservescreditforthis.Heevenguessedtheexistenceof
transmembraneproteins.
In1966,theunit-membrane paradigmcameunderattack
(111)
becauseitwasbelievedthatthestructureofcellmem-
branesmust bemorecomplicatedthanthetheoryseemedto
imply.Inpointing to thefactthatallmembraneshadthesame
kindofbasicstructural plan,theimpressionwasgiventhatall
membranesweremolecularlyidentical.Thiswas, ofcourse,a
complete misunderstanding(96).Theunit-membranemodel
wasincompleteinthatitdid not deal withmembranefluidity
norwiththe idea ofpenetratingproteins.Itwasdeficientin
that
itimpliedthatmembraneproteinswereunfoldedin the
samemannerasproteinsatair-waterinterfaces.However,its
majorfeatures-theuniversalityofthelipidbilayerandchem-
icalasymmetry-aregenerallyacceptedtoday.
THE SUBUNITM ODE LS:
Variousalternativemodelsin
1925
THEIOURNALOfCELLBIOLOGY"VOLUME91,1981
whichthebilayerwasalteredorinterruptedinavarietyof
wayswereproposedinthe1960sbySjbstrand(112,113),
LenardandSinger(114),Greenandhiscolleagues(115,116),
andBenson(117,118).Theessenceofthesemodelswasthat
thebilayerwasnot thedominantstructure,butthatlipid
moleculeswerearrangedin variouspatternsinthemembrane.
TheBensonmodelwasthemostextreme;themembrane
consistedofathinlayerofproteinwithlipidmoleculessimply
intercalatedinavarietyofways.Thepresent-day Sjbstrand
andBarajasmodels(119-121)aresomewhatsimilar.Theseall
fall
moreorlessintothe generalrubricof"subunit" models,
andtheearlieronesweredealtwithquitethoroughlyina
reviewin1969byStoeckeniusandEngleman(14),inwhich
theyconcludedthatthebilayermodelwastheonlyreasonable
one.Thebasicfactherewasthatitwasfoundimpossibleto
breakupanymembranestructureintosubunitsofuniform
compositionthatwouldreassembleintoafunctionalmem-
brane.Another problemwasthat thisconcept impliedthat
membranebiogenesisoccurredbyadditionsofaliquotsof
componentsasaggregatesofmanymoleculesthatservedas
buildingblocksforthewholestructure.Studiesofmembrane
biogenesisshouldthenhaveshownevidenceofparallelin-
creasesofatleastsomesetofrelatedcomponents.Nosuch
increaseswerefound.Forexample,Siekevitz,Palade,andco-
workers(122-128)showedthatindividualnascentmembrane
proteinsareinsertedintodevelopingmembranesystemsat
differenttimes,andthatproteinsinmembranesturn overat
differentrates,independentofoneanotherandindependent
oftheturnoverofmembranephospholipids.
TheModernPeriod
FREEZE-FRACTURE-ETCH(FIFE)ELECTRONMICROS-
CoP Y:Freeze-fracture-etchelectronmicroscopyhasbecome
importantinstructuralstudiesofmembranes.Steere
introducedtheFFEtechniquein1957(129),anditwasdevel-
opedfurtherin 1961byMoorandMuhlethaler(130),who
notedthatmembranesfractured transversely givethesame
triple-layeredunitmembraneappearancefoundinsections.
Theybelievedthatwhenthefractureplanefollowedthe plane
of
themembraneitranalongitssurface.Brantonandhis
associateslater(131-133)proposedthatfrozenbiologicalmem-
branestendtofracturecentrally,andnotedthatthefractured
replicatedsurfacesdisplayedparticlesabout50-100thindi-
ameter(Fig.4);daSilvaandBranton(134)provedthatthe
particleswerelocatedinside theerythrocytemembraneby
labelingtheexternalsurfaceswithferritinandidentifyingitin
theexternal etch(ES)facesinaplaneoutsidetheparticulate
fracture
faces.Theparticlesweremuchmore numerousonthe
protoplasmic(PF)thanontheexternalfracture(EF)facesof
erythrocytemembranes.Almostallfracturedmembranes
showedthisdistributionofparticles,buttheywerenotusually
observedinthenervemyelinsheathandinpurelipidmodel
systems (135-140).Pitsinthecomplementaryfracturefaces,
althoughsometimesseen,wereusuallyabsent.Generally,ret-
inalrodouter segmentmembranesalsofailed toshowdiscrete
particles(141-145).
THEFLUIDMOSAIC(FM)MODEL:
Singer(11,12)
andSingerandNicholsonin1972(146)proposedanew
membrane
paradigmwhichtheycalledthe"fluidmosaic"
model.Thisretainedthebilayerconcept,butintroducedanew
way
of lookingatthedistributionofprotein.Boththeouter
andinnersurfacesweredepicted aslargelynakedlipid.The
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
FIGURE4
Freeze-fracturemicrographofportionsof severalhumanerythrocytemembranes.Thelargeareaofconcavemembrane
totheleftcenterrepresentstheexternalfracture(EF)face ofanerythrocytemembrane.Theconvexfracturefacetotheright
representstheprotoplasmicfracture(PF)faceofanothererythrocytemembrane .Notetheparticles50-100Aindiameterthat are
scattered
irregularlyallover bothkindsoffracturefaces.ThesearemorenumerousonthePF facethanontheEFface .Micrograph
fromH.P.Beall.x47,500.
proteinwasvisualizedasmacromoleculesembeddedinthe
bilayerinaniceberglikefashion,penetratingeitherhalforall
thewaythrough(Fig.5).Theproteinmoleculestraversingthe
bilayerwerevisualizedbySinger(12) ashavingwater-filled
holesintheircenterthatsubservedmembranetransportfunc-
tions.Theemphasisinthismodelwason anextremedegreeof
fluidity,basedontheworkofFryeandEdidin(147),demon-
stratingfluidity inmembranesbyafluorescentdye-labeling
technique.Theproteinmoleculeswerevisualizedasbeing
completely
free totranslate laterallyintheliquidbilayer.The
modelofferedareadyexplanationforthepresenceof50-100A
intramembraneparticles(IMPS)inFFEpreparations,andit
rapidlybecamethegenerallyacceptedmembraneparadigm.
Somefeaturesof theoriginalmodelneedtoberevised.For
example,Singerhasrecognizedtheinadequacyofdepicting
thebilayeras avirtuallynakedstructure.Ifmembranesinvivo
weregenerallynakedlipid bilayerswithoutcontinuouslayers
ofproteinoneithersurface,theywouldhavethemechano-
chemicalpropertiesoflipidbilayers.Evans(148)andLaCelle
(149)independentlycomparedthemechanochemicalproper-
tiesofseveraldifferentkinds ofcellmembraneswithlipid
bilayersandfoundthemtoberadically different.In1974,
Singer(12,150)proposedthat spectrin(151)madeameshwork
onthecytoplasmicsurfaceof the redcellmembraneand
restrictedthemotionofintegralproteins.Thisaddedfeature
madetheFMmodelcompatiblewiththemechanochemical
properties.
Butanimportantproblemstillremained,becauseeventoday
themodelcallsfor theexternalsurfacesofmembranestobe
mostlynakedlipidbilayers.Conceivably,insomespecialcases
likethepurplemembrane(2)withcloselypackedtransmem-
FIGURE5
Diagramofthefluid-mosaicmodelofmembranestruc-
turefromSingerandNicholson(146).Thebilayerisrepresented
here
bycirclesfortheheadgroupwithtwolinesforeachhydro-
phobictail.Proteinisrepresentedbythecross-hatchedparticles
embeddedinthebilayer.
braneproteins,thelipidisnakedinpatchesinvivo,butthere
isnofirmevidenceforthisandcertainlynobasisforgeneral-
izingsuchafeature.Totake theerythrocyteasanexample,if
thelipidwerenakedexternally,onewouldexpecttosee
essentiallynodifferencesinthesusceptibilityofintactred
bloodcells,ghosts,orlipidbilayerstophospholipases.Thisis
notthecase.ZwaalandRoelofson(152)reportedthatsome
phospholipasesare activeonintacterythrocytemembranes,
somehavelittleeffect,butallare activeonghosts.Ottolenghi
(153)hasprepareda highlypurifiedphospholipaseAZand
foundnoeffect atallonintacthumanerythrocytes,although
ghostswereattackedreadily.AdamichandDennis(154)found
ROBERTSON
MembraneStructure
193s
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
Freeze-fracturemicrographofportionsof severalhumanerythrocytemembranes.Thelargeareaofconcavemembrane
totheleftcenterrepresentstheexternalfracture(EF)face ofanerythrocytemembrane.Theconvexfracturefacetotheright
representstheprotoplasmicfracture(PF)faceofanothererythrocytemembrane .Notetheparticles50-100Aindiameterthat are
scattered
irregularlyallover bothkindsoffracturefaces.ThesearemorenumerousonthePF facethanontheEFface .Micrograph
fromH.P.Beall.x47,500.
proteinwasvisualizedasmacromoleculesembeddedinthe
bilayerinaniceberglikefashion,penetratingeitherhalforall
thewaythrough(Fig.5).Theproteinmoleculestraversingthe
bilayerwerevisualizedbySinger(12) ashavingwater-filled
holesintheircenterthatsubservedmembranetransportfunc-
tions.Theemphasisinthismodelwason anextremedegreeof
fluidity,basedontheworkofFryeandEdidin(147),demon-
stratingfluidity inmembranesbyafluorescentdye-labeling
technique.Theproteinmoleculeswerevisualizedasbeing
completely
free totranslate laterallyintheliquidbilayer.The
modelofferedareadyexplanationforthepresenceof50-100A
intramembraneparticles(IMPS)inFFEpreparations,andit
rapidlybecamethegenerallyacceptedmembraneparadigm.
Somefeaturesof theoriginalmodelneedtoberevised.For
example,Singerhasrecognizedtheinadequacyofdepicting
thebilayeras avirtuallynakedstructure.Ifmembranesinvivo
weregenerallynakedlipid bilayerswithoutcontinuouslayers
ofproteinoneithersurface,theywouldhavethemechano-
chemicalpropertiesoflipidbilayers.Evans(148)andLaCelle
(149)independentlycomparedthemechanochemicalproper-
tiesofseveraldifferentkinds ofcellmembraneswithlipid
bilayersandfoundthemtoberadically different.In1974,
Singer(12,150)proposedthat spectrin(151)madeameshwork
onthecytoplasmicsurfaceof the redcellmembraneand
restrictedthemotionofintegralproteins.Thisaddedfeature
madetheFMmodelcompatiblewiththemechanochemical
properties.
Butanimportantproblemstillremained,becauseeventoday
themodelcallsfor theexternalsurfacesofmembranestobe
mostlynakedlipidbilayers.Conceivably,insomespecialcases
likethepurplemembrane(2)withcloselypackedtransmem-
FIGURE5
Diagramofthefluid-mosaicmodelofmembranestruc-
turefromSingerandNicholson(146).Thebilayerisrepresented
here
bycirclesfortheheadgroupwithtwolinesforeachhydro-
phobictail.Proteinisrepresentedbythecross-hatchedparticles
embeddedinthebilayer.
braneproteins,thelipidisnakedinpatchesinvivo,butthere
isnofirmevidenceforthisandcertainlynobasisforgeneral-
izingsuchafeature.Totake theerythrocyteasanexample,if
thelipidwerenakedexternally,onewouldexpecttosee
essentiallynodifferencesinthesusceptibilityofintactred
bloodcells,ghosts,orlipidbilayerstophospholipases.Thisis
notthecase.ZwaalandRoelofson(152)reportedthatsome
phospholipasesare activeonintacterythrocytemembranes,
somehavelittleeffect,butallare activeonghosts.Ottolenghi
(153)hasprepareda highlypurifiedphospholipaseAZand
foundnoeffect atallonintacthumanerythrocytes,although
ghostswereattackedreadily.AdamichandDennis(154)found
ROBERTSON
MembraneStructure
193s
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
thatlessthan1°Ioof thephospholipidsinintacterythrocytes
werehydrolyzedbyphospholipaseA2,whereas38%ofthetotal
phospholipidsofghostswerehydrolysedunderthesamecon-
ditions.Itseemsclear,then,thattheintacterythrocytemem-
braneisdefinitelymoreresistanttophospholipasesthanare
ghostsorlipidmicelles.Thelipidaccessibilityismuchlessthan
theFMmodelimplies.
Weshallnowturntoamoredetailedconsiderationofthe
currentlyacceptedconcepts ofmembranestructure.
PRESENTCONCEPTSOFTHEMOLECULAR
ORGANIZATIONOFMEMBRANES
General
Itisnowgenerallyagreedthatallbiologicalmembranes
containalipidbilayer,as describedabove.Theprotein:lipid:
carbohydrateratiosvaryconsiderably byweightfrommem-
branetomembrane,rangingfrom 75:25:0withthepurple
membraneofHalobacteriumhalobiumatoneextreme,through
49:43:8forhumanerythrocytes,to18:79:3formyelinatthe
otherextreme(6).ThelipidsaremainlyPC,phosphatidyl
ethanolamine(PE),phosphatidylserine(PS),sphingomyelin,
andcholesterol.Somemembranesarehighinglycolipids,
phosphotidylinositol,orcardiolipin.Onethingthatallthe
lipidshaveincommonisamphiphilicity(52).
Therearetwokinds ofmembraneproteins:peripheraland
integral(12).Theformerareoperationallydefinedasones
easilyremovablebyionicmanipulationsandthelatterbythe
needfordetergentsorotherchaotropicagents,becausethey
arehydrophobicallybonded.Someareconfinedtoonesideof
the
bilayer(ecto-orendo-[7]),andsomepenetrateitpartially
orcompletely.Thedominantmassofthisintegralproteinin
mostmembranesislocatedinthepolarregions of thebilayer.
Forexample,inarecentneutrondiffractionstudyofretinal
rodsYeageretal.(155)estimatedthatthetotalmassofthe
rhodopsinmoleculethatcanbeintheanhydroushydrocarbon
regionis15-20%.Theband-3proteinoferythrocytemem-
braneshasonly19%ofitsmassinthepenetratingcomponent
(156).Thebacteriorhodopsinmoleculeisanexception;more
thanhalfitsmassisinthehydrocarbonregion(2,157,158).
Theoperationaldefinitionofperipheralandintegralmem-
braneproteinsdoesnotholdstrictly,asSingernotedin1974
(12).Forexample,ligatin(159-163),an-10,000dmembrane-
bindingprotein,althoughhydrophobicallybondedtolipidand
henceintegralbythiscriterion,canberemovedby10-40mM
Ca",takingwithitacomplementoflipid,mainlytriphos-
phoinositolplussomePCandcholesterol.It isa highlynega-
tivelychargedglycoprotein,whichdoesnothaveahighcom-
plementofhydrophobicaminoacids.Exactlyhowitisbound
isnotclear,butitresidesin theexternal surfaceofcertain
membranes,whereitfunctionstobindcertainectoproteins.
TheErythrocyteMembraneasanExample
Despiteitsspecialization,theerythrocytemembranehas
beenmorewidelystudiedsince1971thananyothermembrane.
It isabout60:40protein:lipidbyweightandcontainsatleast
adozenwell-definedproteins(10,35,164-170).Theseare
generallyreferredtobythenumbersof thepositionsthey
assumeaselectrophoreticbandsinpolyacrylamidegels,follow-
ingtheterminologyofFairbankset al.(164).Chemicallabeling
experiments,firstbyBretscherin1971(171-173)andlater
moredefinitivelybyWhiteleyandBerg(174),andproteolytic
1945
THEJOURNALOFCELLBIOLOGY"VOLUME91,1981
dissectionexperimentsfirstdonein1971-1972byStecket al.
(10,175)andothers(156,176,177)haveresultedinthelocation
ofmostoftheseproteinsasperipheralorintegral(seereference
35forbibliography).Bands1,2and4-6areperipheralendo-
proteins.Bands1and2represent spectrin(10,35,178-180),
alsocalledtektinA(181,182);band4isuncharacterized;band
5isactin(10,35);andband6isglyceraldehyde-3-phosphate
dehydrogenase(35, 183,184).Actinandspectrinareassociated
(35,185).Periodicacid-Schiff(PAS)positive1and2andband
3are themajorglycoproteins.PAS1and2areinterconvertible
sialoglycoproteins
(35)identifiedwithglycophorinA,ablood-
groupsubstance(186)thatmakesup75%ofthisgroupal-
thoughonly-2%ofthetotalprotein(187).Followingearlier
workbyWinzlerin1969(188)andMorawieckiin1964(189),
Marchesiandhiscolleagues(190-192)studiedthisprotein
extensively.Itsaminoacidsequencehasbeendetermined(35,
192).Itcontains astretchof 20hydrophobicaminoacidsthat
transversesthelipidbilayer.Incommonwithalltheglycopro-
teins,thecarbohydrate moietyisexternal(35).Band3 isa
transmembraneglycoproteincontainingnosialicacidthat
makesup20-25%of thetotalmembraneprotein(169).It
functionsinaniontransport(169)andcanbechemically
crosslinkedwithspectrin(193).Anumberofotherproteins,
suchasNa+K`ouabain-sensitiveadenosinetriphosphatase
(ATPase)(194,195)andacetylcholineesterase(AChe)are
presentinlesseramounts(35).Theformerisbelievedtobea
transmembraneprotein.Thelatterisexternallylocated(196).
TheIntramembraneParticle
IMPSareclearlyassociatedwithproteinsinmembranes.In
1971,forexample,Branton(133)foundthatthenumberof
IMPsarereducedinredcellmembranestreatedwithproteases.
PurelipidbilayersdonotnormallycontainIMPs,although
theymaydisplaysomepatternedsubstructureundersome
conditions(197).Hongand Hubbel(198)showedin1972that
additionofrhodopsintobilayervesiclescause theappearance
ofIMPs.However,in1975DeamerandYamanaca(199)found
thatsarcoplasmicreticulummembranestreatedwithproteases
totheextentthatalltheirproteincomponentswere reducedto
polypeptidefragmentsof10,000dorlessstillcontained about
thesamenumberofIMPs,althoughthey losttheirdominant
PFfaceorientation.Verkleijet al.(200)haveshownthata
puremixedlipidsystemmay,inthepresenceofCa",display
typical-100AIMPswithcorrespondingpitsinthecomple-
mentaryfracture faces.Attemptshave beenmadetocorrelate
the
numbersofIMPswith theknownnumbersofcopiesof
certainintegralproteins.Althoughneverexact,suchcorrela-
tions
sometimeshaveappearedtobeclose(10)butinother
casesnocorrelationatallwasfound(201).Twogroups-da
Silvaet al.in1971(202)andTillacketal.in1972(203)-
showed
thattherewasarelationshipbetweentheexternal
surface
componentsofglycophorinandband-3proteinin
erythrocytesandtheintramembraneparticlesseeninFFE
preparations.daSilvain 1972(204),ElgsaegerandBrantonin
1974
(205),andothersmorerecently(206,207)havepresented
evidence ofIMPaggregationphenomena.Thesearealtered
whenspectrinisextractedfromerythrocyteghosts,whichis
interpretedasindicationsofthetransmembraneconnections
ofglycophorinandband3 withspectrin.TheworkofTilney
andDetmers(185)suggestshowspectrin,actin,andthegly-
coproteinsmightinteract,andit isbelieved(208)thatunder
thecontrolofan endogenouskinaseandphosphatasethese
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
werehydrolyzedbyphospholipaseA2,whereas38%ofthetotal
phospholipidsofghostswerehydrolysedunderthesamecon-
ditions.Itseemsclear,then,thattheintacterythrocytemem-
braneisdefinitelymoreresistanttophospholipasesthanare
ghostsorlipidmicelles.Thelipidaccessibilityismuchlessthan
theFMmodelimplies.
Weshallnowturntoamoredetailedconsiderationofthe
currentlyacceptedconcepts ofmembranestructure.
PRESENTCONCEPTSOFTHEMOLECULAR
ORGANIZATIONOFMEMBRANES
General
Itisnowgenerallyagreedthatallbiologicalmembranes
containalipidbilayer,as describedabove.Theprotein:lipid:
carbohydrateratiosvaryconsiderably byweightfrommem-
branetomembrane,rangingfrom 75:25:0withthepurple
membraneofHalobacteriumhalobiumatoneextreme,through
49:43:8forhumanerythrocytes,to18:79:3formyelinatthe
otherextreme(6).ThelipidsaremainlyPC,phosphatidyl
ethanolamine(PE),phosphatidylserine(PS),sphingomyelin,
andcholesterol.Somemembranesarehighinglycolipids,
phosphotidylinositol,orcardiolipin.Onethingthatallthe
lipidshaveincommonisamphiphilicity(52).
Therearetwokinds ofmembraneproteins:peripheraland
integral(12).Theformerareoperationallydefinedasones
easilyremovablebyionicmanipulationsandthelatterbythe
needfordetergentsorotherchaotropicagents,becausethey
arehydrophobicallybonded.Someareconfinedtoonesideof
the
bilayer(ecto-orendo-[7]),andsomepenetrateitpartially
orcompletely.Thedominantmassofthisintegralproteinin
mostmembranesislocatedinthepolarregions of thebilayer.
Forexample,inarecentneutrondiffractionstudyofretinal
rodsYeageretal.(155)estimatedthatthetotalmassofthe
rhodopsinmoleculethatcanbeintheanhydroushydrocarbon
regionis15-20%.Theband-3proteinoferythrocytemem-
braneshasonly19%ofitsmassinthepenetratingcomponent
(156).Thebacteriorhodopsinmoleculeisanexception;more
thanhalfitsmassisinthehydrocarbonregion(2,157,158).
Theoperationaldefinitionofperipheralandintegralmem-
braneproteinsdoesnotholdstrictly,asSingernotedin1974
(12).Forexample,ligatin(159-163),an-10,000dmembrane-
bindingprotein,althoughhydrophobicallybondedtolipidand
henceintegralbythiscriterion,canberemovedby10-40mM
Ca",takingwithitacomplementoflipid,mainlytriphos-
phoinositolplussomePCandcholesterol.It isa highlynega-
tivelychargedglycoprotein,whichdoesnothaveahighcom-
plementofhydrophobicaminoacids.Exactlyhowitisbound
isnotclear,butitresidesin theexternal surfaceofcertain
membranes,whereitfunctionstobindcertainectoproteins.
TheErythrocyteMembraneasanExample
Despiteitsspecialization,theerythrocytemembranehas
beenmorewidelystudiedsince1971thananyothermembrane.
It isabout60:40protein:lipidbyweightandcontainsatleast
adozenwell-definedproteins(10,35,164-170).Theseare
generallyreferredtobythenumbersof thepositionsthey
assumeaselectrophoreticbandsinpolyacrylamidegels,follow-
ingtheterminologyofFairbankset al.(164).Chemicallabeling
experiments,firstbyBretscherin1971(171-173)andlater
moredefinitivelybyWhiteleyandBerg(174),andproteolytic
1945
THEJOURNALOFCELLBIOLOGY"VOLUME91,1981
dissectionexperimentsfirstdonein1971-1972byStecket al.
(10,175)andothers(156,176,177)haveresultedinthelocation
ofmostoftheseproteinsasperipheralorintegral(seereference
35forbibliography).Bands1,2and4-6areperipheralendo-
proteins.Bands1and2represent spectrin(10,35,178-180),
alsocalledtektinA(181,182);band4isuncharacterized;band
5isactin(10,35);andband6isglyceraldehyde-3-phosphate
dehydrogenase(35, 183,184).Actinandspectrinareassociated
(35,185).Periodicacid-Schiff(PAS)positive1and2andband
3are themajorglycoproteins.PAS1and2areinterconvertible
sialoglycoproteins
(35)identifiedwithglycophorinA,ablood-
groupsubstance(186)thatmakesup75%ofthisgroupal-
thoughonly-2%ofthetotalprotein(187).Followingearlier
workbyWinzlerin1969(188)andMorawieckiin1964(189),
Marchesiandhiscolleagues(190-192)studiedthisprotein
extensively.Itsaminoacidsequencehasbeendetermined(35,
192).Itcontains astretchof 20hydrophobicaminoacidsthat
transversesthelipidbilayer.Incommonwithalltheglycopro-
teins,thecarbohydrate moietyisexternal(35).Band3 isa
transmembraneglycoproteincontainingnosialicacidthat
makesup20-25%of thetotalmembraneprotein(169).It
functionsinaniontransport(169)andcanbechemically
crosslinkedwithspectrin(193).Anumberofotherproteins,
suchasNa+K`ouabain-sensitiveadenosinetriphosphatase
(ATPase)(194,195)andacetylcholineesterase(AChe)are
presentinlesseramounts(35).Theformerisbelievedtobea
transmembraneprotein.Thelatterisexternallylocated(196).
TheIntramembraneParticle
IMPSareclearlyassociatedwithproteinsinmembranes.In
1971,forexample,Branton(133)foundthatthenumberof
IMPsarereducedinredcellmembranestreatedwithproteases.
PurelipidbilayersdonotnormallycontainIMPs,although
theymaydisplaysomepatternedsubstructureundersome
conditions(197).Hongand Hubbel(198)showedin1972that
additionofrhodopsintobilayervesiclescause theappearance
ofIMPs.However,in1975DeamerandYamanaca(199)found
thatsarcoplasmicreticulummembranestreatedwithproteases
totheextentthatalltheirproteincomponentswere reducedto
polypeptidefragmentsof10,000dorlessstillcontained about
thesamenumberofIMPs,althoughthey losttheirdominant
PFfaceorientation.Verkleijet al.(200)haveshownthata
puremixedlipidsystemmay,inthepresenceofCa",display
typical-100AIMPswithcorrespondingpitsinthecomple-
mentaryfracture faces.Attemptshave beenmadetocorrelate
the
numbersofIMPswith theknownnumbersofcopiesof
certainintegralproteins.Althoughneverexact,suchcorrela-
tions
sometimeshaveappearedtobeclose(10)butinother
casesnocorrelationatallwasfound(201).Twogroups-da
Silvaet al.in1971(202)andTillacketal.in1972(203)-
showed
thattherewasarelationshipbetweentheexternal
surface
componentsofglycophorinandband-3proteinin
erythrocytesandtheintramembraneparticlesseeninFFE
preparations.daSilvain 1972(204),ElgsaegerandBrantonin
1974
(205),andothersmorerecently(206,207)havepresented
evidence ofIMPaggregationphenomena.Thesearealtered
whenspectrinisextractedfromerythrocyteghosts,whichis
interpretedasindicationsofthetransmembraneconnections
ofglycophorinandband3 withspectrin.TheworkofTilney
andDetmers(185)suggestshowspectrin,actin,andthegly-
coproteinsmightinteract,andit isbelieved(208)thatunder
thecontrolofan endogenouskinaseandphosphatasethese
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
proteinsinteractwithadenosinetriphosphatetoregulatethe
shapeof thecell.
Weinsteinetal.(156)studiederythrocytemembranes
treated,beforefracturing,by mildproteolyticdigestionto
removesomeexternal proteinandleaving theband-3protein
intact.Surfaceprojectionswereseen in theESfaceafter
etchingrelatedtounderlying-66AIMPSinthePFfaces.
Theywereinterpretedasthesurface38,000dcomponentsof
theband-3protein.Inside-out erythrocytemembranevesicles
depleted ofspectrinandactinwith theband-3proteinintact
showed,inetchedpreparations,granulofibrillarcomponents
havinganaveragediameterof 90A.Proteasedigestionunder
conditionsthatresulted inreleaseofthe40,000dcomponent
resultedin lossofthesegranulofibrillarcomponents.Thepat-
temofdispositionandrelativenumbersofthesecomponents
beforeremovalwasconsistentwiththeirbeingthe40,000d
cytoplasmicsurfacecomponentoftheband-3protein.Inthat
thewholemoleculehasamolecularweightof95,000d,the
transmembranecomponentis17,000dinweight.Evenasa
dimerplus theglycophorinchain,thisistoo smalltocorre-
spondexactlytothe
-
66AIMPs.
GlycophorinAdoesnotappeartobeimportantinthe
productionofIMPs,becauseit isabsentinararebloodtype
En(a-)
(186, 209),andthereisnoeffectontheIMPsexceptfor
changesin thedynamicaggregationpropertiesthatsuggest
associationofthecytoplasmiccomponentofglycophorinwith
spectrin.
Becauseof theknownassociationofsomeIMPswithprotein
thereisatendencyto regardanyIMPliterallyasametal-
plated protein
molecule.Thisoversimplificationhasledto
muchconfusionintheliterature.TheexactnatureofIMPsis
notyetresolved.
TheErythrocyteLipids
Ithasbeenknownsince1971thatthedistributionof the
lipidconstituentsofthebilayerinerythrocytesisasymmetric
(154,171-173, 210-216),withaminolipidslocatedprimarily
in the
internalmonolayerandcholineandsphingolipids
localizedmainlyin theexternalmonolayer.Thiswasfirst
suggestedbylabelingexperimentsconductedin1971by
Bretscher(171-173),whoshowedthattherelativelyimper-
meantagentFMMPdoesnotreactwiththeaminophospho-
glyceridesinintactcells,whereasbothPSandPEreactinopen
ghosts.Theseexperimentsweresomewhatinconclusivebecause
theydependedontheassumptionthatnomajormolecular
rearrangements occur
in ghosts.However,GordeskyandMar-
inetti(211)foundin1973thattrinitro-benzenesulfonate,a
nonpenetratingreagent,did notlabelPSandonlypartially
labeledPEin intactcells,thus agreeing withBretscher'scon-
clusions.VanDeenen(204)hasreviewedtheevidencefrom
selectivedegradationof thephospholipidsofintacterythrocyte
andghostmembranesrelatingto thisproblem(215, 216),and
AdamichandDennis(154)havereportedsimilarfindingswith
cobravenomphospholipaseA2.Theenzymaticdegradation
anddouble-labelingexperimentsagreeingeneral.Presumably,
glycolipidsarealsolocalizedin the outermonolayerbecause
therearenosugarresiduesonthecytoplasmicsurface.
Bretscher(210)pointedout in 1973thesignificanceofob-
servationsofRouseret al.(217)onthefattyacidcomposition
ofthephospholipidofhumanerythrocytesinrelationtothe
above.TheaminolipidsPEandPScontainmuchmore20:4
andtotalpolyunsaturatedfattyacidsthandothecholine-
containingphospholipids.Sphingomyelincontainsmuch16:0,
24:0,and24:1fattyacids.PSishighestin18:0acids.Thus,the
inner halfofthebilayercontainsmoreunsaturatedlipids,
whereasthe outerhalfcontainsmoresaturatedandlonger
chain
lipids.
Molecularspectroscopicstudiesin 1971byKornbergand
McConnell
(218)haveshownthatlipidsdonotbecometrans-
locatedspontaneouslyfromonesideof amodelbilayertothe
other(flip-flop)withinatime spanofhours,but theymaydo
so fairlyfrequentlywithin atimespanofminutesinexcitable
membranesbyspecialunknownmechanismsmediatedby
protein.Incontrast,theseauthors(219)showedthatlateral
diffusioninPCbilayersiseight ordersofmagnitudemore
rapid.Van Deenan(214)notedthattheflip-floprateofphos-
phatidyl
choline,whichisvirtuallyundetectableinbilayers,is
speededupbytheaddition ofglycophorin.
FLUIDITY
Lipid
Onekindoffluidityinvolvesmotionof thecarbonchain
relative totheheadgroup;cholesterol influencesthisgreatly.
Thereissomeevidencethat cholesterolislocalizedto the outer
half of
themyelinmembrane(108)andtheerythrocytemem-
brane(220,221).Ithasbeenshown(222-224)that cholesterol
hasacondensingeffectonphospholipidsinmonolayersor
bilayers,decreasing theaveragearea perlipidmolecule.This
impliesthatitmakesthemembranemorerigid.However,it
alsohasthe functionofconvertinglipidsinthe stiff,extended
(Ls )conformationtothemoreliquid(La )conformation(139,
140),afunctionitshareswithdoublebondsinthelipidcarbon
chainsandhigher temperature.Thisisduetotheproduction
ofpotentialspacesinthecenterofthebilayerthatresultfrom
cholesterolbeingshorterthanmostmembranelipids.
Thetermfluidityisusedinanothersense, inwhich whole
lipidmoleculesdiffuselaterally.Anincreaseinthelocalized
mobilityoftheindividualhydrocarbonchainsisnotnecessarily
implicated.Stillanotherkindoffluidityinvolvesrotationof
lipidmolecules.Specialtechniquesbeyondthescopeofthis
articleareusedtodetecteachofthesevariouskindsoffluidity.
LipidandLipid-ProteinDomainFluidity
Instillanotherkindoffluidity,aggregatesof lipidmolecules
maydiffuseasunitseitheras freedomainsortogetherwith
proteinconstituentswithwhichtheyarespecificallyassociated.
Theileummembraneofsucklingrats(160-163)isagood
exampleoflipid-proteindomainfluidity.
Thereisalsoevidencethatseparatelipiddomainsmayexist
inmembraneswithoutnecessarilybeingassociatedwithpro-
tein.In1978,MarinettiandCrain(213),usingpenetratingand
nonpenetratingcrosslinkingprobes,providedgoodevidence
fortheasymmetricdistributionofphospholipidsinerythrocyte
membranesas wellasformosaicassociationsofgroupsof
specificphospholipidsandspecificphospholipid-proteincom-
plexes.ShimshickandMcConnel(225)foundin1973thatif
twolipidsdifferinginchain lengthbyatleasttwo carbon
atomsaremixedandkeptatatemperatureabovethephase
transitionofoneandbelowthatoftheother,thelipidsseparate
intotwophases,oneintheLoandtheother theL,#state.Ranck
et al.(226)andCostelloandGulik-Krzywicki(197)showed
thatdifferentlipiddomainsproducedifferenttexturesinfrac-
turefacesinFFEpreparations.Klausneretal.in1980(227)
usedfluorescentprobestudiestoproduceevidencethatlym-
ROBERTSON
MembraneStructure
195s
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
shapeof thecell.
Weinsteinetal.(156)studiederythrocytemembranes
treated,beforefracturing,by mildproteolyticdigestionto
removesomeexternal proteinandleaving theband-3protein
intact.Surfaceprojectionswereseen in theESfaceafter
etchingrelatedtounderlying-66AIMPSinthePFfaces.
Theywereinterpretedasthesurface38,000dcomponentsof
theband-3protein.Inside-out erythrocytemembranevesicles
depleted ofspectrinandactinwith theband-3proteinintact
showed,inetchedpreparations,granulofibrillarcomponents
havinganaveragediameterof 90A.Proteasedigestionunder
conditionsthatresulted inreleaseofthe40,000dcomponent
resultedin lossofthesegranulofibrillarcomponents.Thepat-
temofdispositionandrelativenumbersofthesecomponents
beforeremovalwasconsistentwiththeirbeingthe40,000d
cytoplasmicsurfacecomponentoftheband-3protein.Inthat
thewholemoleculehasamolecularweightof95,000d,the
transmembranecomponentis17,000dinweight.Evenasa
dimerplus theglycophorinchain,thisistoo smalltocorre-
spondexactlytothe
-
66AIMPs.
GlycophorinAdoesnotappeartobeimportantinthe
productionofIMPs,becauseit isabsentinararebloodtype
En(a-)
(186, 209),andthereisnoeffectontheIMPsexceptfor
changesin thedynamicaggregationpropertiesthatsuggest
associationofthecytoplasmiccomponentofglycophorinwith
spectrin.
Becauseof theknownassociationofsomeIMPswithprotein
thereisatendencyto regardanyIMPliterallyasametal-
plated protein
molecule.Thisoversimplificationhasledto
muchconfusionintheliterature.TheexactnatureofIMPsis
notyetresolved.
TheErythrocyteLipids
Ithasbeenknownsince1971thatthedistributionof the
lipidconstituentsofthebilayerinerythrocytesisasymmetric
(154,171-173, 210-216),withaminolipidslocatedprimarily
in the
internalmonolayerandcholineandsphingolipids
localizedmainlyin theexternalmonolayer.Thiswasfirst
suggestedbylabelingexperimentsconductedin1971by
Bretscher(171-173),whoshowedthattherelativelyimper-
meantagentFMMPdoesnotreactwiththeaminophospho-
glyceridesinintactcells,whereasbothPSandPEreactinopen
ghosts.Theseexperimentsweresomewhatinconclusivebecause
theydependedontheassumptionthatnomajormolecular
rearrangements occur
in ghosts.However,GordeskyandMar-
inetti(211)foundin1973thattrinitro-benzenesulfonate,a
nonpenetratingreagent,did notlabelPSandonlypartially
labeledPEin intactcells,thus agreeing withBretscher'scon-
clusions.VanDeenen(204)hasreviewedtheevidencefrom
selectivedegradationof thephospholipidsofintacterythrocyte
andghostmembranesrelatingto thisproblem(215, 216),and
AdamichandDennis(154)havereportedsimilarfindingswith
cobravenomphospholipaseA2.Theenzymaticdegradation
anddouble-labelingexperimentsagreeingeneral.Presumably,
glycolipidsarealsolocalizedin the outermonolayerbecause
therearenosugarresiduesonthecytoplasmicsurface.
Bretscher(210)pointedout in 1973thesignificanceofob-
servationsofRouseret al.(217)onthefattyacidcomposition
ofthephospholipidofhumanerythrocytesinrelationtothe
above.TheaminolipidsPEandPScontainmuchmore20:4
andtotalpolyunsaturatedfattyacidsthandothecholine-
containingphospholipids.Sphingomyelincontainsmuch16:0,
24:0,and24:1fattyacids.PSishighestin18:0acids.Thus,the
inner halfofthebilayercontainsmoreunsaturatedlipids,
whereasthe outerhalfcontainsmoresaturatedandlonger
chain
lipids.
Molecularspectroscopicstudiesin 1971byKornbergand
McConnell
(218)haveshownthatlipidsdonotbecometrans-
locatedspontaneouslyfromonesideof amodelbilayertothe
other(flip-flop)withinatime spanofhours,but theymaydo
so fairlyfrequentlywithin atimespanofminutesinexcitable
membranesbyspecialunknownmechanismsmediatedby
protein.Incontrast,theseauthors(219)showedthatlateral
diffusioninPCbilayersiseight ordersofmagnitudemore
rapid.Van Deenan(214)notedthattheflip-floprateofphos-
phatidyl
choline,whichisvirtuallyundetectableinbilayers,is
speededupbytheaddition ofglycophorin.
FLUIDITY
Lipid
Onekindoffluidityinvolvesmotionof thecarbonchain
relative totheheadgroup;cholesterol influencesthisgreatly.
Thereissomeevidencethat cholesterolislocalizedto the outer
half of
themyelinmembrane(108)andtheerythrocytemem-
brane(220,221).Ithasbeenshown(222-224)that cholesterol
hasacondensingeffectonphospholipidsinmonolayersor
bilayers,decreasing theaveragearea perlipidmolecule.This
impliesthatitmakesthemembranemorerigid.However,it
alsohasthe functionofconvertinglipidsinthe stiff,extended
(Ls )conformationtothemoreliquid(La )conformation(139,
140),afunctionitshareswithdoublebondsinthelipidcarbon
chainsandhigher temperature.Thisisduetotheproduction
ofpotentialspacesinthecenterofthebilayerthatresultfrom
cholesterolbeingshorterthanmostmembranelipids.
Thetermfluidityisusedinanothersense, inwhich whole
lipidmoleculesdiffuselaterally.Anincreaseinthelocalized
mobilityoftheindividualhydrocarbonchainsisnotnecessarily
implicated.Stillanotherkindoffluidityinvolvesrotationof
lipidmolecules.Specialtechniquesbeyondthescopeofthis
articleareusedtodetecteachofthesevariouskindsoffluidity.
LipidandLipid-ProteinDomainFluidity
Instillanotherkindoffluidity,aggregatesof lipidmolecules
maydiffuseasunitseitheras freedomainsortogetherwith
proteinconstituentswithwhichtheyarespecificallyassociated.
Theileummembraneofsucklingrats(160-163)isagood
exampleoflipid-proteindomainfluidity.
Thereisalsoevidencethatseparatelipiddomainsmayexist
inmembraneswithoutnecessarilybeingassociatedwithpro-
tein.In1978,MarinettiandCrain(213),usingpenetratingand
nonpenetratingcrosslinkingprobes,providedgoodevidence
fortheasymmetricdistributionofphospholipidsinerythrocyte
membranesas wellasformosaicassociationsofgroupsof
specificphospholipidsandspecificphospholipid-proteincom-
plexes.ShimshickandMcConnel(225)foundin1973thatif
twolipidsdifferinginchain lengthbyatleasttwo carbon
atomsaremixedandkeptatatemperatureabovethephase
transitionofoneandbelowthatoftheother,thelipidsseparate
intotwophases,oneintheLoandtheother theL,#state.Ranck
et al.(226)andCostelloandGulik-Krzywicki(197)showed
thatdifferentlipiddomainsproducedifferenttexturesinfrac-
turefacesinFFEpreparations.Klausneretal.in1980(227)
usedfluorescentprobestudiestoproduceevidencethatlym-
ROBERTSON
MembraneStructure
195s
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
phocytemembranescontainseparatelipiddomains.FFEstud-
iesofurinarybladderepithelialcellmembranes(228-231)also
suggestthatsuchlipiddomainsexistandthattheymaybe
relatedtoproteinaswellastotheproductionofartifactual
Imps
.
Thespreadingandmixingoffluorescentsurfacemarkers
firstdescribedin1970by FryeandEdiden(147),aswellas
studiesofcappingphenomena(232-237)ofsurfacemarkers
suchasferritinorfluorescentlylabeledlectinsorantibodies
representakindofmosaicfluiditythatprobablyinvolvesboth
lipidsandproteins.
THE PURPLE MEMBRANE :
ThepurplemembraneofH.
halobiumhasbeenstudiedextensively since1968byStoeck-
eniuset al
.(2,238-243).Thishighlydifferentiatedmembrane,
whichpumpsprotonsfromthecell(241,243),appearsin
patchesintheplasmamembraneofH
.halobiumthatcontain
a purpleproteinpigmentmoleculeof 26,000dcalledbacteri-
orhodopsin(BR)(240),andunusualphospholipids(244)
.Un-
win andHenderson(157-159)showedthatBRisarrangedin
thelipidbilayercoreasatransmembraneproteininahexag-
onallatticewithP
3spacegroupsymmetryandalatticeconstant
of62A.Sevenalphahelicestraversethelipidbilayerforeach
molecule.
Theisolatedpurplemembraneprobablyshouldberegarded
as a highlyspecializedresidualskeletalmembranefromwhich
peripheralproteinshave beenremovedduringisolation,be-
causethereisnoevidencethatitexists invivo as anaked
bilayer.Similarly,ReynoldsandTrayer(245)foundthateryth-
rocytemembranescouldbestrippedofupto90%oftheir
proteinbytreatment withdiluteionicsolutionscontaining
EGTA,producingdegradederythrocytemembranes.Thus,in
termsofgeneralmembranemodels,theisolatedpurplemem-
Outside
GIU
SEA~S"PRO,
VAL
Glr33
MIT
ALA39
tY5
«u
TON
67
THR
NIFT
It(
VAL
GIN
PRO
Au
PigGtr
ASN
GIN75
GlY
A B C
19
6s
THEJOURNAL OFCELLBIOLOGY "VOLUME 91,1981
r
D
braneisrelevantonlyinthesenseof beinganextremeexample
ofconcentrationofhydrophobictransmembraneproteininthe
bilayer.
FisherandStoeckenius(246)inaFFEstudyoftheisolated
purplemembranenotedtheabsenceoftheusualglobular
IMPs.TheyobservedmuchsmallerparticlesinthePFfaces.
Theyregardedtheseasaggregatesof 9-12BRmolecules.
KueblerandGross(247)andUsukuraet al.(248)reported a
latticeof-50A-diameterparticlesinthePFfaces.Robertson
etal
.(249)haveobservedsimilarparticlesandidentifiedthem
asaggregatesofthreeBRmoleculesdisplayedmainlyby
decoration
.Thusthepurplemembranecontainsatransmem-
braneproteinwithtransmembranemasscomparabletothatof
othertransmembraneproteins,buttheindividualmolecules
havenotbeenresolved.
Engelmanetal.(250)derivedatheoreticalmolecularmodel
(Fig.6) ofBRinsituprimarilyfromtheknownaminoacid
sequencedeterminedbyOvchinikovetal.(251;seealso252
and253).Themodelcontainsnine chargedresiduesburiedin
thehydrophobiccore,butatleastsixareneutralized.BRisan
inside-outprotein,since thereisa higher concentration of
hydrophobicresiduesnexttothelipid,with the hydrophilic
residuestendingtobemoreconcentrated intheinteriorofthe
molecule(254,255).However,thepartofthemoleculeembed-
dedinthehydrocarbonregionisdominantlyhydrophobic.If
onecountsthehydrophobicandhydrophilicresiduesdrawnin
Fig.6within thehydrocarbonregionofthebilayer,73%are
hydrophobic.Thecytoplasmicsurfacestratumis50%hydro-
phobicandtheoutsidesurfacestratumis58%hydrophobic.
Theaveragehydrophobicity[H
o(ave)] asdefinedbyBigelow
(256)is1613forthecoreand728and754for thecytoplasmic
andexternal surfacestrata
.Thereisthus astronggradient of
SIR
TNA
TYR
SER
AU
ALA
PHI
r
TNR
YR
VAL
ALA
vAL
u
r
ALA
SER
VALA
VAl
VAL
VAL
PRO
lapLIU
TYR
Top
VAL
ILE
LEU
GLY
SER192
GU
GV
IEU
LIU
ILE
QY
LIU
GLY
AU
U
PH<
VAL
I
SEA
VAL
1
VA
WWI
l
ALA
lE
ILI
ASN
GAY06
U
LEU
PRO
VAL
lu
FIGURE 6ModeltakenfromEnglemanetal.(250)ofthearrangementofthepolypeptideofbacteriorhodopsinacrossthe
membraneofHalobacteriumhalobium.Theseven alphahelicesarelabeledA-Gstartingfromtheaminoterminus.Thehatch
marksindicatetheapproximatelocation ofthelipidhydrocarbonregionsOandOindicate positiveandnegativecharges
respectively.ThesequenceistakenfromOvchinikovetal.(251).
Inside ~/
TIIR
~
,
VAIN='IV
oG1
A
PHf
au
»R
lfU
TYR
ALA
ILt
THR
THR
lfU
VAL
~U
lIU
AU
lfU
AS
IfU
IlE
lEU
AU
IlF
VAL
GY
GY
lEU
VAl
TNRGl`
Liu
Go
MCT
LIU
VAL
PRO
ALA
Ilf
ALA
lIU
LIU
LIU
PRO
TNR
TNR
AU
AS
GY
lIU
R!7
lu
LIU
ILE
TYR
W
Au
ILIA
TNRG,,
LIUALA
lfU
RP
Prf
TNRRAT
TYR
LEU
SIR
LI
T
PNt
lEU
TRP
AS
AU
TYR
TRIPAU
~RTYR
GY
TINT
GY
LEU
(AV
VAL
TNR
SIR
1~
Top
vAL
IIPRO~I/
LYLI
GIY
IE
YAII
~
TNR
TYR©
" --
1
-Ii -
PRO
ILE
AUno No
U RrT
ALA
SER
AIA
~7! Au
PRO
PROtbl SIR
GY $IN
GU
A
QGV ALA
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
iesofurinarybladderepithelialcellmembranes(228-231)also
suggestthatsuchlipiddomainsexistandthattheymaybe
relatedtoproteinaswellastotheproductionofartifactual
Imps
.
Thespreadingandmixingoffluorescentsurfacemarkers
firstdescribedin1970by FryeandEdiden(147),aswellas
studiesofcappingphenomena(232-237)ofsurfacemarkers
suchasferritinorfluorescentlylabeledlectinsorantibodies
representakindofmosaicfluiditythatprobablyinvolvesboth
lipidsandproteins.
THE PURPLE MEMBRANE :
ThepurplemembraneofH.
halobiumhasbeenstudiedextensively since1968byStoeck-
eniuset al
.(2,238-243).Thishighlydifferentiatedmembrane,
whichpumpsprotonsfromthecell(241,243),appearsin
patchesintheplasmamembraneofH
.halobiumthatcontain
a purpleproteinpigmentmoleculeof 26,000dcalledbacteri-
orhodopsin(BR)(240),andunusualphospholipids(244)
.Un-
win andHenderson(157-159)showedthatBRisarrangedin
thelipidbilayercoreasatransmembraneproteininahexag-
onallatticewithP
3spacegroupsymmetryandalatticeconstant
of62A.Sevenalphahelicestraversethelipidbilayerforeach
molecule.
Theisolatedpurplemembraneprobablyshouldberegarded
as a highlyspecializedresidualskeletalmembranefromwhich
peripheralproteinshave beenremovedduringisolation,be-
causethereisnoevidencethatitexists invivo as anaked
bilayer.Similarly,ReynoldsandTrayer(245)foundthateryth-
rocytemembranescouldbestrippedofupto90%oftheir
proteinbytreatment withdiluteionicsolutionscontaining
EGTA,producingdegradederythrocytemembranes.Thus,in
termsofgeneralmembranemodels,theisolatedpurplemem-
Outside
GIU
SEA~S"PRO,
VAL
Glr33
MIT
ALA39
tY5
«u
TON
67
THR
NIFT
It(
VAL
GIN
PRO
Au
PigGtr
ASN
GIN75
GlY
A B C
19
6s
THEJOURNAL OFCELLBIOLOGY "VOLUME 91,1981
r
D
braneisrelevantonlyinthesenseof beinganextremeexample
ofconcentrationofhydrophobictransmembraneproteininthe
bilayer.
FisherandStoeckenius(246)inaFFEstudyoftheisolated
purplemembranenotedtheabsenceoftheusualglobular
IMPs.TheyobservedmuchsmallerparticlesinthePFfaces.
Theyregardedtheseasaggregatesof 9-12BRmolecules.
KueblerandGross(247)andUsukuraet al.(248)reported a
latticeof-50A-diameterparticlesinthePFfaces.Robertson
etal
.(249)haveobservedsimilarparticlesandidentifiedthem
asaggregatesofthreeBRmoleculesdisplayedmainlyby
decoration
.Thusthepurplemembranecontainsatransmem-
braneproteinwithtransmembranemasscomparabletothatof
othertransmembraneproteins,buttheindividualmolecules
havenotbeenresolved.
Engelmanetal.(250)derivedatheoreticalmolecularmodel
(Fig.6) ofBRinsituprimarilyfromtheknownaminoacid
sequencedeterminedbyOvchinikovetal.(251;seealso252
and253).Themodelcontainsnine chargedresiduesburiedin
thehydrophobiccore,butatleastsixareneutralized.BRisan
inside-outprotein,since thereisa higher concentration of
hydrophobicresiduesnexttothelipid,with the hydrophilic
residuestendingtobemoreconcentrated intheinteriorofthe
molecule(254,255).However,thepartofthemoleculeembed-
dedinthehydrocarbonregionisdominantlyhydrophobic.If
onecountsthehydrophobicandhydrophilicresiduesdrawnin
Fig.6within thehydrocarbonregionofthebilayer,73%are
hydrophobic.Thecytoplasmicsurfacestratumis50%hydro-
phobicandtheoutsidesurfacestratumis58%hydrophobic.
Theaveragehydrophobicity[H
o(ave)] asdefinedbyBigelow
(256)is1613forthecoreand728and754for thecytoplasmic
andexternal surfacestrata
.Thereisthus astronggradient of
SIR
TNA
TYR
SER
AU
ALA
PHI
r
TNR
YR
VAL
ALA
vAL
u
r
ALA
SER
VALA
VAl
VAL
VAL
PRO
lapLIU
TYR
Top
VAL
ILE
LEU
GLY
SER192
GU
GV
IEU
LIU
ILE
QY
LIU
GLY
AU
U
PH<
VAL
I
SEA
VAL
1
VA
WWI
l
ALA
lE
ILI
ASN
GAY06
U
LEU
PRO
VAL
lu
FIGURE 6ModeltakenfromEnglemanetal.(250)ofthearrangementofthepolypeptideofbacteriorhodopsinacrossthe
membraneofHalobacteriumhalobium.Theseven alphahelicesarelabeledA-Gstartingfromtheaminoterminus.Thehatch
marksindicatetheapproximatelocation ofthelipidhydrocarbonregionsOandOindicate positiveandnegativecharges
respectively.ThesequenceistakenfromOvchinikovetal.(251).
Inside ~/
TIIR
~
,
VAIN='IV
oG1
A
PHf
au
»R
lfU
TYR
ALA
ILt
THR
THR
lfU
VAL
~U
lIU
AU
lfU
AS
IfU
IlE
lEU
AU
IlF
VAL
GY
GY
lEU
VAl
TNRGl`
Liu
Go
MCT
LIU
VAL
PRO
ALA
Ilf
ALA
lIU
LIU
LIU
PRO
TNR
TNR
AU
AS
GY
lIU
R!7
lu
LIU
ILE
TYR
W
Au
ILIA
TNRG,,
LIUALA
lfU
RP
Prf
TNRRAT
TYR
LEU
SIR
LI
T
PNt
lEU
TRP
AS
AU
TYR
TRIPAU
~RTYR
GY
TINT
GY
LEU
(AV
VAL
TNR
SIR
1~
Top
vAL
IIPRO~I/
LYLI
GIY
IE
YAII
~
TNR
TYR©
" --
1
-Ii -
PRO
ILE
AUno No
U RrT
ALA
SER
AIA
~7! Au
PRO
PROtbl SIR
GY $IN
GU
A
QGV ALA
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
hydrophobicitytowardthecenterof themembrane.Thisex-
plainsthedistributionofheavy-metalstaininthesectioned
membraneinwhichtheproteinisnotstained(5,249).Accord-
ingtoZaccaiandEngleman(254),whostudiedthemembrane
byneutrondiffraction,thereisnosuggestionofa hydrophilic
channelthatcouldsupportacolumnofwateracrossthe
membranethroughthemolecule.Theystatedthatnopockets
intheproteincontain 12 ormorewatermolecules,andthat
theirresultsexcludethepossibilitythatpassivetransportoccurs
viaabulkwaterchannelintheprotein.
Weshallnowturntothegapjunction,the study ofwhich
hasledtoamuchbetterunderstandingofmembranestructure.
THEG A P)UNCTION:In theearly1950s Sj6strandetal.
(257-259)obtainedsomeofthefirstelectronmicrographsof
sectionsofepithelialcellsinwhichsurfacemembranescould
beseen.Theysawintercellularboundariesasdenselines<100
Athicknexttocytoplasmseparatedbyaclearinterzone
-160
.~wide(258).Theyproposedthattheclearzonesrepresented
lipidandthedensezonesprotein,basedonSjostrand'ssections
ofnervemyelininwhichtheconstituentmembraneswerefirst
visualized(260),andinwhichhecorrectlyinterpretedthe
densestrata torepresentproteinandthelightstratalipid.In
1958,heandhiscolleagues(261)studied cardiacmuscleand
resolvedthepreviouslyobservedsingledenselinesthatborder
theintercellular clearzonesintotwotriple-layeredstructures,
each-75Athick.However,theyinterpreted theseasmono-
layersofprotein.Theyobservednarrowinginthewidthsof
theclear intercellularzonesinsomeplaceswhichwecan
recognizetodayasgapjunctions,but theyevidentlybelieved
thatthesesimplyrepresentedvariationsinthethicknessof the
intercellularlipidlayers .Theyalsoobserveddesmosomes,and
heretherewassomematerialbetweenthemembranesthat
stained.Thisledthemtopostulatethat thedesmosomewas
thesiteofelectrotoniccoupling.Thisisinterestinghistorically
becauseitillustratesparticularlywell theimportanceof the
paradigminthedevelopmentofthefield.Theunit-membrane
conceptclearedupthismatterbyidentifyingeachofthetriple-
layeredstructuresseenat intercellularboundariesasacomplete
cellmembrane,including alipidbilayerandtheclear-100-
150
Agapbetweenthemembranesas highlyhydratedextra-
cellularspacethatcouldbevariedinthicknessexperimentally.
Aboutthesametime(in1959)itwasdemonstratedthatthe
gapsnormallypresentbetweentheSchwanncellmembrane
andtheaxonmembraneintheinternodalregionsofmyelinated
nervefiberswereclosedinthe juxta-terminalmyelinatedregion
atnodesofRanvier(262).Thegapswerepresentbetweenthe
Schwanncellnodalprocessesandtheaxonmembrane.Itwas
recognizedthatthegapclosuresatthenodewouldfunctionto
preventlateralflow of ionsalongthesurfaceof the axon,thus
facilitatingsaltatoryconduction.Theclosureoftheintercellu-
largapsinthemyelinsheathwasalsonecessaryfor saltatory
conductioninthesamesense.Theterm"externalcompound
membrane"
(92)wasproposedfortwosuchmembranesin
closecontact.Thedevelopmentoftheaboveconceptslaidthe
basisforthe understandinginthenextdecadeofthefunctions
ofboththeoccluding junctionandthegapjunction.
Closecontactsofmembraneswerefirstreportedinthe
crayfishmedian-giant-to-motorsynapses in 1953(263),but
werenot understood.Later,in1961,micrographsofthemem-
branesin thissynapseshowedtheunitmembraneswithcom-
pleteclosureofthesynapticcleft(264).FurshpanandPotter
(265)
showedin1959thatthissynapsewasanelectricalrecti-
fyingone,andthesignificanceoftheclosureofthecleftas a
possiblemorphologicalbasisforelectroniccouplingwasim-
mediately apparent.KarrerandCoxin1960(266)described
membranecontactsinintercalateddisksincardiac-typemuscle
inmousethoracicandlungveins,andreferredtothemas
"externalcompoundmembranes."Theyrecognizedtheirprob-
ablefunction assitesof transmissionofexcitationbetween
musclecells.Theywerethusthefirsttopublishclearelectron
micrographsofthin sectionsofwhatarenowcalledgap
junctionsandtodeducetheirfunctioncorrectly.
In1962,FurshpanandFurakawa(267)foundevidencein
theMauthnercellofthegoldfishmedullaofthepresenceof
electricalsynapses,andFurshpanpresented evidence(268)that
theseweretheclubendingsofBartelmez(269)onthelateral
dendrite.Thinsectionsoftheseendingsshowedmembrane
contacts-0.3-0.5pmindiameterineach clubending.In
frontalview,thesejunctionsshowedahexagonalarrayof
subunits with alatticeconstant of80-90Awhichhadnot
beenseen before(Figs.7and8).Thesecontactswerecalled
"synapticdisks"(270-272).Figure9(273)isaFFEmicrograph
ofsucha junction.DeweyandBarr(274,275)independently
foundevidenceofelectricalcontactsbetweensmoothand
cardiacmusclefibersanddiscovered astructureverysimilarto
thesynapticdiskthattheyrelatedtoelectricaltransmission,
although,likeKarrer andCox, theydid notobservesubstruc-
tureinthemembranes.Theycalled thesecontacts"nexuses."
In1965,BennedettiandEmmelot(276)observedmembranes
withpatternslikethesynapticdiskinnegativelystainedmem-
branefractionsisolatedfromliver.Latertheyidentifiedthese
astightoroccludingjunctions(277).
FarquharandPaladein1963(278)proposedasetofterms
for thevariousjunctions seenin epithelialtissues.Inearlier
workwithsectionsofintestinalepithelium(Fig.34inreference
96)andinmesaxonsofmyelinated nervefibers(Fig.34in
reference264),regionsoffocalpartialfusionwerenotedin
whichunitmembranescameintoclosecontactandtheoverall
thicknesswasreducedtowellbelowtwicethethicknessofone
unitmembrane.Atthesametime,thetwoexternaldensestrata
disappearedfocallyandthelightcentralzonesmergedfor
distancesof~100Aorso .Thesewereregardedaszonesof
focalpartialmolecularfusion.FarquharandPalade(278-280),
insurveying variousepitheliainsections,sawthesestructures
FIGURE7Transversesectionofgapjunctionfrom aclubending
onthelateraldendriteoftheMauthnercell in agoldfishbrain.The
two
membranesarecloselyapposedand a beadingisseeninthe
region ofcontactrepeatingina period ofabout 80A.Thisprepa-
rationwasfixedinpotassiumpermanganateandembeddedin
Araldite.Undertheseconditionsthegapinthejunctiondoesnow
show up.Thisstructurewascalleda"synapticdisc"intheoriginal
publicationinwhichitwaspresentedin1963(271).x443,500.
ROBERTSON
MembraneStructure
197s
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
plainsthedistributionofheavy-metalstaininthesectioned
membraneinwhichtheproteinisnotstained(5,249).Accord-
ingtoZaccaiandEngleman(254),whostudiedthemembrane
byneutrondiffraction,thereisnosuggestionofa hydrophilic
channelthatcouldsupportacolumnofwateracrossthe
membranethroughthemolecule.Theystatedthatnopockets
intheproteincontain 12 ormorewatermolecules,andthat
theirresultsexcludethepossibilitythatpassivetransportoccurs
viaabulkwaterchannelintheprotein.
Weshallnowturntothegapjunction,the study ofwhich
hasledtoamuchbetterunderstandingofmembranestructure.
THEG A P)UNCTION:In theearly1950s Sj6strandetal.
(257-259)obtainedsomeofthefirstelectronmicrographsof
sectionsofepithelialcellsinwhichsurfacemembranescould
beseen.Theysawintercellularboundariesasdenselines<100
Athicknexttocytoplasmseparatedbyaclearinterzone
-160
.~wide(258).Theyproposedthattheclearzonesrepresented
lipidandthedensezonesprotein,basedonSjostrand'ssections
ofnervemyelininwhichtheconstituentmembraneswerefirst
visualized(260),andinwhichhecorrectlyinterpretedthe
densestrata torepresentproteinandthelightstratalipid.In
1958,heandhiscolleagues(261)studied cardiacmuscleand
resolvedthepreviouslyobservedsingledenselinesthatborder
theintercellular clearzonesintotwotriple-layeredstructures,
each-75Athick.However,theyinterpreted theseasmono-
layersofprotein.Theyobservednarrowinginthewidthsof
theclear intercellularzonesinsomeplaceswhichwecan
recognizetodayasgapjunctions,but theyevidentlybelieved
thatthesesimplyrepresentedvariationsinthethicknessof the
intercellularlipidlayers .Theyalsoobserveddesmosomes,and
heretherewassomematerialbetweenthemembranesthat
stained.Thisledthemtopostulatethat thedesmosomewas
thesiteofelectrotoniccoupling.Thisisinterestinghistorically
becauseitillustratesparticularlywell theimportanceof the
paradigminthedevelopmentofthefield.Theunit-membrane
conceptclearedupthismatterbyidentifyingeachofthetriple-
layeredstructuresseenat intercellularboundariesasacomplete
cellmembrane,including alipidbilayerandtheclear-100-
150
Agapbetweenthemembranesas highlyhydratedextra-
cellularspacethatcouldbevariedinthicknessexperimentally.
Aboutthesametime(in1959)itwasdemonstratedthatthe
gapsnormallypresentbetweentheSchwanncellmembrane
andtheaxonmembraneintheinternodalregionsofmyelinated
nervefiberswereclosedinthe juxta-terminalmyelinatedregion
atnodesofRanvier(262).Thegapswerepresentbetweenthe
Schwanncellnodalprocessesandtheaxonmembrane.Itwas
recognizedthatthegapclosuresatthenodewouldfunctionto
preventlateralflow of ionsalongthesurfaceof the axon,thus
facilitatingsaltatoryconduction.Theclosureoftheintercellu-
largapsinthemyelinsheathwasalsonecessaryfor saltatory
conductioninthesamesense.Theterm"externalcompound
membrane"
(92)wasproposedfortwosuchmembranesin
closecontact.Thedevelopmentoftheaboveconceptslaidthe
basisforthe understandinginthenextdecadeofthefunctions
ofboththeoccluding junctionandthegapjunction.
Closecontactsofmembraneswerefirstreportedinthe
crayfishmedian-giant-to-motorsynapses in 1953(263),but
werenot understood.Later,in1961,micrographsofthemem-
branesin thissynapseshowedtheunitmembraneswithcom-
pleteclosureofthesynapticcleft(264).FurshpanandPotter
(265)
showedin1959thatthissynapsewasanelectricalrecti-
fyingone,andthesignificanceoftheclosureofthecleftas a
possiblemorphologicalbasisforelectroniccouplingwasim-
mediately apparent.KarrerandCoxin1960(266)described
membranecontactsinintercalateddisksincardiac-typemuscle
inmousethoracicandlungveins,andreferredtothemas
"externalcompoundmembranes."Theyrecognizedtheirprob-
ablefunction assitesof transmissionofexcitationbetween
musclecells.Theywerethusthefirsttopublishclearelectron
micrographsofthin sectionsofwhatarenowcalledgap
junctionsandtodeducetheirfunctioncorrectly.
In1962,FurshpanandFurakawa(267)foundevidencein
theMauthnercellofthegoldfishmedullaofthepresenceof
electricalsynapses,andFurshpanpresented evidence(268)that
theseweretheclubendingsofBartelmez(269)onthelateral
dendrite.Thinsectionsoftheseendingsshowedmembrane
contacts-0.3-0.5pmindiameterineach clubending.In
frontalview,thesejunctionsshowedahexagonalarrayof
subunits with alatticeconstant of80-90Awhichhadnot
beenseen before(Figs.7and8).Thesecontactswerecalled
"synapticdisks"(270-272).Figure9(273)isaFFEmicrograph
ofsucha junction.DeweyandBarr(274,275)independently
foundevidenceofelectricalcontactsbetweensmoothand
cardiacmusclefibersanddiscovered astructureverysimilarto
thesynapticdiskthattheyrelatedtoelectricaltransmission,
although,likeKarrer andCox, theydid notobservesubstruc-
tureinthemembranes.Theycalled thesecontacts"nexuses."
In1965,BennedettiandEmmelot(276)observedmembranes
withpatternslikethesynapticdiskinnegativelystainedmem-
branefractionsisolatedfromliver.Latertheyidentifiedthese
astightoroccludingjunctions(277).
FarquharandPaladein1963(278)proposedasetofterms
for thevariousjunctions seenin epithelialtissues.Inearlier
workwithsectionsofintestinalepithelium(Fig.34inreference
96)andinmesaxonsofmyelinated nervefibers(Fig.34in
reference264),regionsoffocalpartialfusionwerenotedin
whichunitmembranescameintoclosecontactandtheoverall
thicknesswasreducedtowellbelowtwicethethicknessofone
unitmembrane.Atthesametime,thetwoexternaldensestrata
disappearedfocallyandthelightcentralzonesmergedfor
distancesof~100Aorso .Thesewereregardedaszonesof
focalpartialmolecularfusion.FarquharandPalade(278-280),
insurveying variousepitheliainsections,sawthesestructures
FIGURE7Transversesectionofgapjunctionfrom aclubending
onthelateraldendriteoftheMauthnercell in agoldfishbrain.The
two
membranesarecloselyapposedand a beadingisseeninthe
region ofcontactrepeatingina period ofabout 80A.Thisprepa-
rationwasfixedinpotassiumpermanganateandembeddedin
Araldite.Undertheseconditionsthegapinthejunctiondoesnow
show up.Thisstructurewascalleda"synapticdisc"intheoriginal
publicationinwhichitwaspresentedin1963(271).x443,500.
ROBERTSON
MembraneStructure
197s
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
FIGURE8FrontalViewofagapjunctionsimilartotheonepre-
sentedinFig.7reproduced fromthesamepaper(271).Notethata
fairlyregularhexagonalarrayoffacetswasseen.Eachfacethasa
denseborderabout25Awidesurroundingaclearzoneinthe
center
ofwhichisaspotabout20-25Aindiameter.x134,000.
andcorrectlydeducedtheirfunctioninlimitinglateraldiffu-
sionofmaterialbetweencells.Theyvisualizedthemasbelthke
structuresaroundthejunctionsofepithelialcellsdesignedto
limitpassageofmaterialbetweencells.Theycalledthem
"zonulaoccludens,"but alsousedthe olderterm"tightjunc-
tion."Theynotedthepunctateregionsofpartialmembrane
fusion,andreferredtothemasmembrane"pinches."Figure
10isaFFEmicrographofanoccluding junction(281).The
ridgescorrespondtothepunctateregionsofpartialfusionin
sections.FarquharandPalade(278)emphasizedtheclose
contactof theinterveningmembranes, whichtheycalledre-
gionsofmembrane"fusion,"usingfusioninthesense ofclose
apposition.Theyincludedthe"nexus"(274)inthesameclass
asthe"tight"junction.Theydistinguishedtwoothertypesof
junction,thezonulaadhaerensandthemaculaadhaerens,to
bothofwhichtheyassigned anattachmentfunction.The
zonulaeoccludensandmaculaadhaerenshadbeendescribed
bylightmicroscopistsandcalledrespectively"terminalbars"
and"desmosomes."
Amatterrelatedtothe evolution ofthisfieldisthe use of
lanthanumas atracer.Lettvinetal.(282)notedthatLa...
acted
intheperipheralnervoussystemlikea"superCa++."W.
F.PickardsynthesizedLa(Mn04)3andDoggenweilerand
Frenk(283)useditasafixativein1965 .Theynotedthat
La+++,eitherintroducedinthiswayoraddedbyincubationin
La(NO3)3beforefixation,impartedgreatdensityinthe inter-
cellularsubstancesofthenervoussystem.RevelandKamowski
(284)in1967thendevelopedanextracellulartracertechnique
basedonthiswork,bycombininglanthanumsaltswithglutar-
aldehyde.Extracellularspaceswerestained generallyandthe
techniquesshowedupregionsofclosecontactbetweenepithe-
lialcellsinavarietyofdifferenttissues whichresembledvery
muchthesynapticdiscafterKMn04fixation.Inorder to
1985
THEJOURNALOFCELLBIOLOGY"VOLUME91,1981
distinguishthesejunctions sharplyfromthe"tight"junctions
or
occludingjunctions describedearlierby FarquharandPa-
lade(278-280),they applied tothemtheterm"gap junction."
Thetermisnowalmostuniversallyuseddespitethe suggestion
bySiminoescuetal.(285)thatthesejunctionsbecalled"mac-
ulaecommunicantes,"orcommunicatingjunctions.
In1968,Kreutziger(286)producedthefirstelectronmicro-
graphsofFFEpreparations ofgapjunctions.Heobservedon
onefracturefaceparticlesinaroughlyhexagonalarraywitha
center-to-centerspacingof80-90Aand,onthe other,a
correspondingpatternofpits(seeFig.9).Unfortunately,he
misidentifiedthefracturefacesandplacedtheparticlesinthe
EFface.Thiswascarriedonbyothers(273,287, 288)until
ChalcroftandBullivant(289)in 1970andSteereandSommer
(290)in 1972independentlyproducedcomplementarydouble
replicasandcorrectlyidentifiedthefracturefaces.It isnow
generallyagreedthatinvertebratestheparticlesarealwaysin
thePFfaces.McNuttandWeinsteinin1970(291)presented
amodelshowingtransversechannelscrossingthejunctional
membranesineachface.
Peracchia(292,293) hasshownthatinthecrayfishlateral
giantseptalgapjunctionstheIMPlocalizationinthefracture
facesisreversed; i.e.,theparticlesareintheEFfacesandpits
inthePFfaces.Theparticles,thoughinroughhexagonal
array, arespacedabouttwiceasfarapart(-200A)andthe
gapisdistinctlywider(40-50A).PeracchiaandDulhunty
(294)alsofoundthattheparticlesaremuchmoretightlyand
regularlyarrayed(150-155 Aspacing)ifthejunctionsare
uncoupledbytreatmentwithCa"-andMg'-freesolutions
withEDTA,orbydinitrophenol.
Peracchiaextendedthesestudiestoratgapjunctions(295)
andagainfoundthatcoupledjunctionshavelooser,lessregular
particlearraysspacedat-103-105A,whereasuncoupled ones
havemoretightlypackedparticlesspacedat-85A.
Gapjunctionsareinvolvedinintercellularcommunication
inmanydifferentanimalsandtissues.Alargeliteraturehas
developedaroundcombinedmicroelectrodeandstructralstud-
iesofintercellularcommunication byuseoffluorescentdyes
andothersubstances,datingbacktoearlystudiesbyLowen-
steinfrom1966onwards(296),Potteret al.(297),Pappasand
Bennett(298),Sheridan(299),andothers.Itisbeyondthe
scope of
thisarticletoreviewthisliterature,butarticlesby
Lowenstein(300)andWarner(301)inarecentsymposium
volumemaybe consultedforkeyreferences.
Severalgroups,followingthe pioneeringworkofBenedetti
andEmmelot,isolatedgapjunctionsfromvarious sourcesfor
detailedchemicalandstructuralanalysis.Goodenoughand
Stoeckenius(302)reported amethodusingcollagenaseand
hyaluronidasedigestionaftertreatmentwiththe detergent
sarcosyl,which,asEvansandGurd(303)alsofound,selectively
dissolvednonjunctionalmembranes.Theliteratureisconfusing
inthatanumberofdifferentmolecular-weightproteinswere
isolatedasfollows:byGoodenoughin1974(304)-34,000d,
18,000d,andadoubletat10,000dcalledconnexinAandB;
byGilulain1974(305)-10,000dand20,000d;byDuniaet
al.in1974(306)-34,000d,13,000d,and26,000d;byGood-
enoughin 1976(307)-9,000dand18,000d,thelattercalled
"connexin"andusedasthebasisforamodel(308);byDuguid
andRevelin 1975(309)-26,000dand36,000d;byBenedetti
et al.in1976(310)-34,000dand26,000d;byCulvenarand
Evansin1977(311)-38,000dand40,000d;byZampighiand
Robertsonin1977(312)andZampighiin1978(313)-25,000
d;byEhrhartandChauveauin1977(314)-34,000d;by
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
sentedinFig.7reproduced fromthesamepaper(271).Notethata
fairlyregularhexagonalarrayoffacetswasseen.Eachfacethasa
denseborderabout25Awidesurroundingaclearzoneinthe
center
ofwhichisaspotabout20-25Aindiameter.x134,000.
andcorrectlydeducedtheirfunctioninlimitinglateraldiffu-
sionofmaterialbetweencells.Theyvisualizedthemasbelthke
structuresaroundthejunctionsofepithelialcellsdesignedto
limitpassageofmaterialbetweencells.Theycalledthem
"zonulaoccludens,"but alsousedthe olderterm"tightjunc-
tion."Theynotedthepunctateregionsofpartialmembrane
fusion,andreferredtothemasmembrane"pinches."Figure
10isaFFEmicrographofanoccluding junction(281).The
ridgescorrespondtothepunctateregionsofpartialfusionin
sections.FarquharandPalade(278)emphasizedtheclose
contactof theinterveningmembranes, whichtheycalledre-
gionsofmembrane"fusion,"usingfusioninthesense ofclose
apposition.Theyincludedthe"nexus"(274)inthesameclass
asthe"tight"junction.Theydistinguishedtwoothertypesof
junction,thezonulaadhaerensandthemaculaadhaerens,to
bothofwhichtheyassigned anattachmentfunction.The
zonulaeoccludensandmaculaadhaerenshadbeendescribed
bylightmicroscopistsandcalledrespectively"terminalbars"
and"desmosomes."
Amatterrelatedtothe evolution ofthisfieldisthe use of
lanthanumas atracer.Lettvinetal.(282)notedthatLa...
acted
intheperipheralnervoussystemlikea"superCa++."W.
F.PickardsynthesizedLa(Mn04)3andDoggenweilerand
Frenk(283)useditasafixativein1965 .Theynotedthat
La+++,eitherintroducedinthiswayoraddedbyincubationin
La(NO3)3beforefixation,impartedgreatdensityinthe inter-
cellularsubstancesofthenervoussystem.RevelandKamowski
(284)in1967thendevelopedanextracellulartracertechnique
basedonthiswork,bycombininglanthanumsaltswithglutar-
aldehyde.Extracellularspaceswerestained generallyandthe
techniquesshowedupregionsofclosecontactbetweenepithe-
lialcellsinavarietyofdifferenttissues whichresembledvery
muchthesynapticdiscafterKMn04fixation.Inorder to
1985
THEJOURNALOFCELLBIOLOGY"VOLUME91,1981
distinguishthesejunctions sharplyfromthe"tight"junctions
or
occludingjunctions describedearlierby FarquharandPa-
lade(278-280),they applied tothemtheterm"gap junction."
Thetermisnowalmostuniversallyuseddespitethe suggestion
bySiminoescuetal.(285)thatthesejunctionsbecalled"mac-
ulaecommunicantes,"orcommunicatingjunctions.
In1968,Kreutziger(286)producedthefirstelectronmicro-
graphsofFFEpreparations ofgapjunctions.Heobservedon
onefracturefaceparticlesinaroughlyhexagonalarraywitha
center-to-centerspacingof80-90Aand,onthe other,a
correspondingpatternofpits(seeFig.9).Unfortunately,he
misidentifiedthefracturefacesandplacedtheparticlesinthe
EFface.Thiswascarriedonbyothers(273,287, 288)until
ChalcroftandBullivant(289)in 1970andSteereandSommer
(290)in 1972independentlyproducedcomplementarydouble
replicasandcorrectlyidentifiedthefracturefaces.It isnow
generallyagreedthatinvertebratestheparticlesarealwaysin
thePFfaces.McNuttandWeinsteinin1970(291)presented
amodelshowingtransversechannelscrossingthejunctional
membranesineachface.
Peracchia(292,293) hasshownthatinthecrayfishlateral
giantseptalgapjunctionstheIMPlocalizationinthefracture
facesisreversed; i.e.,theparticlesareintheEFfacesandpits
inthePFfaces.Theparticles,thoughinroughhexagonal
array, arespacedabouttwiceasfarapart(-200A)andthe
gapisdistinctlywider(40-50A).PeracchiaandDulhunty
(294)alsofoundthattheparticlesaremuchmoretightlyand
regularlyarrayed(150-155 Aspacing)ifthejunctionsare
uncoupledbytreatmentwithCa"-andMg'-freesolutions
withEDTA,orbydinitrophenol.
Peracchiaextendedthesestudiestoratgapjunctions(295)
andagainfoundthatcoupledjunctionshavelooser,lessregular
particlearraysspacedat-103-105A,whereasuncoupled ones
havemoretightlypackedparticlesspacedat-85A.
Gapjunctionsareinvolvedinintercellularcommunication
inmanydifferentanimalsandtissues.Alargeliteraturehas
developedaroundcombinedmicroelectrodeandstructralstud-
iesofintercellularcommunication byuseoffluorescentdyes
andothersubstances,datingbacktoearlystudiesbyLowen-
steinfrom1966onwards(296),Potteret al.(297),Pappasand
Bennett(298),Sheridan(299),andothers.Itisbeyondthe
scope of
thisarticletoreviewthisliterature,butarticlesby
Lowenstein(300)andWarner(301)inarecentsymposium
volumemaybe consultedforkeyreferences.
Severalgroups,followingthe pioneeringworkofBenedetti
andEmmelot,isolatedgapjunctionsfromvarious sourcesfor
detailedchemicalandstructuralanalysis.Goodenoughand
Stoeckenius(302)reported amethodusingcollagenaseand
hyaluronidasedigestionaftertreatmentwiththe detergent
sarcosyl,which,asEvansandGurd(303)alsofound,selectively
dissolvednonjunctionalmembranes.Theliteratureisconfusing
inthatanumberofdifferentmolecular-weightproteinswere
isolatedasfollows:byGoodenoughin1974(304)-34,000d,
18,000d,andadoubletat10,000dcalledconnexinAandB;
byGilulain1974(305)-10,000dand20,000d;byDuniaet
al.in1974(306)-34,000d,13,000d,and26,000d;byGood-
enoughin 1976(307)-9,000dand18,000d,thelattercalled
"connexin"andusedasthebasisforamodel(308);byDuguid
andRevelin 1975(309)-26,000dand36,000d;byBenedetti
et al.in1976(310)-34,000dand26,000d;byCulvenarand
Evansin1977(311)-38,000dand40,000d;byZampighiand
Robertsonin1977(312)andZampighiin1978(313)-25,000
d;byEhrhartandChauveauin1977(314)-34,000d;by
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
FIGURE9
Freeze-fracturepreparationof agapjunctiontakenfromGoodenoughandRevel(273).Notethatthejunctionconsists
of ratherirregulararraysofparticlesalternatingwithregionsinwhichpitsareseen .TheparticlesarelocatedonthePF faceofone
ofthejunctionalmembranesandthepitsarelocatedintheEF face.x102,500.
Gilulain1978(315)andHertzbergandGilulain1979(316)-
47,000dand27,000d.
In1979,Hendersonet al.(317)clarifiedsomeofthese
conflictingreports.Theyavoidedenzymetreatmentsforpuri-
fication,used6Murea(310)intheisolationprocedure,and
avoidedboilinginsodiumdodecylsulfate(SDS).Theycon-
cludedthattherewereonly twomolecularspeciespresentat
26,000dand21,000d,andthatthesmallerone wasprobably
adegradationproduct.Thehigher molecular-weightcompo-
nentswereconsideredto resultfromaggregationofthehydro-
phobic26,000dcomponentsinboilingSDS.Thiscomponent
wasreducedto13,000dbytrypsintreatment.Theyperformed
aminoacid analysesonthe26,000dfragmentandnotedthat
ithadahydrophobicityindexdiscriminantfunction,Z =0.322,
referringtotheclassificationof Barrantes(318),whofound
thatintegralmembraneproteinsallhaveZvaluesin excessof-
0.317.Thetrypsin-treatedproteinshowedadistinctincreasein
Zvalueto0.589,duetoa reductioninthetotalcontentof
hydrophilic
aminoacids.Theyconcludedthatthe13,000d
trypsin-resistantcomponentwasprobablyburiedin thelipid
ROBERTSON
MembraneStructure
1995
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
Freeze-fracturepreparationof agapjunctiontakenfromGoodenoughandRevel(273).Notethatthejunctionconsists
of ratherirregulararraysofparticlesalternatingwithregionsinwhichpitsareseen .TheparticlesarelocatedonthePF faceofone
ofthejunctionalmembranesandthepitsarelocatedintheEF face.x102,500.
Gilulain1978(315)andHertzbergandGilulain1979(316)-
47,000dand27,000d.
In1979,Hendersonet al.(317)clarifiedsomeofthese
conflictingreports.Theyavoidedenzymetreatmentsforpuri-
fication,used6Murea(310)intheisolationprocedure,and
avoidedboilinginsodiumdodecylsulfate(SDS).Theycon-
cludedthattherewereonly twomolecularspeciespresentat
26,000dand21,000d,andthatthesmallerone wasprobably
adegradationproduct.Thehigher molecular-weightcompo-
nentswereconsideredto resultfromaggregationofthehydro-
phobic26,000dcomponentsinboilingSDS.Thiscomponent
wasreducedto13,000dbytrypsintreatment.Theyperformed
aminoacid analysesonthe26,000dfragmentandnotedthat
ithadahydrophobicityindexdiscriminantfunction,Z =0.322,
referringtotheclassificationof Barrantes(318),whofound
thatintegralmembraneproteinsallhaveZvaluesin excessof-
0.317.Thetrypsin-treatedproteinshowedadistinctincreasein
Zvalueto0.589,duetoa reductioninthetotalcontentof
hydrophilic
aminoacids.Theyconcludedthatthe13,000d
trypsin-resistantcomponentwasprobablyburiedin thelipid
ROBERTSON
MembraneStructure
1995
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
FIGURE10Freeze-fractureofzonnulaoccludensbetweenglutar-
aldehydefixedepithelial cellsintheileumofanadultrat.After
fixationthejunctionfeaturesacharacteristicbeltlikenetworkof
branchedand anastomosingridges(R)onthePFfaceandcorre-
spondingfurrowsontheEF face.Microvilli(MV)areseenabove.
AnextensivePF faceisseenbelow.x42,380.
asthetransmembranepartofthemolecule.Theyalsonoted
thatthereisa very highmolarratioofcholesterolinthe
junctionsdespitethe detergenttreatment,andsuggestedthatit
playsastructuralrole.
Finbowet al.(319)havereportedanindependentlineof
evidence
whichalsosuggeststhatthemajorgapjunctional
proteinisthe 26,000dcomponent.Thisgrouphadshown(320,
321)thatpartialhepatectomycaused gapjunctionstodisappear
duringapostoperativeperiodof24-28hwithareturnto
normalwithin48h.Theynotedthatthe26,000dcomponent
wasabsentinthe24-28-hperiodandreappearedat48h.There
isnowgeneralagreementthattheprincipalgapjunctional
proteinhasanapparentmolecularweightof26,000d.
Casparetal.(322)andMakowskiet al.(323)conducteda
combinedchemical,electronmicroscope,andX-raydiffraction
study ofisolatedgapjunctionswhichledthemtoproposea
model.Theyreliedheavilyonamicrographpublishedby
Goodenoughin1976(307, 324),interpretedasshowingPTA
fillingan-20Atransversehydrophilicchannelinanisolated
junction.Thisledthemtopostulatetheexistenceofaqueous
-20A-diameter,protein-linedchannelscompletelytraversing
thejunctions.Theysupposedthatflowthroughthesechannels
wasregulatedbyvariationsinthediameterof thechannelin
the regionbetweenthetwomembranes.Theyapplied thename
"connexon"for thecompletechannelsthatrunacrosseach of
thetwomembranesof a given junction.
Ithaslongbeenclearthatsomesortoftransversechannel
structureispresentintheunitmembranesofthe junctions.
Lowenstein,forexample,hasshownthatmoleculesupto20A
indiametercanpassthrough(300).However,electronmicro-
graphsfailed toshowdirectevidenceofanysuch pore.Thus,
ZampighiandRobertson,in1974,(325)foundthatitwas
possibletodegradetheisolatedjunctionselectivelybytreating
itwithEDTAorEGTA.Aftertreatment,the junctionbroke
upintofragmentsthat consistedofonlyafewoftherepeating
units,someofwhichwerefoundtohe ontheirside innegative
stain.Noevidence oftransversechannelswasfound.Achannel
-20Aindiameterwouldbeexpectedtofillandbeseenunder
theseconditions,becausethereisacomparableholeintobacco
mosaicvirusthatfillsreadilywithPTA(326).Itwasconcluded
thatthechannelmustbesmallerthanthe-20Asuggestedby
thesizeofthestainaccumulationbetweenthetwomembranes.
This broughtintofocus theproblemof the nature ofthe
channel.Obviously,onepossibilitywouldbe a tubularstruc-
200s
THEJOURNALOFCELLBIOLOGY"VOLUME91,1981
tureconsisting,perhaps, ofsomethinglikeafl-pleatedsheetof
polypeptidechainsrolledintoacylindricalformwithhydro-
phobicresiduesontheoutsideandhydrophilicresiduesinside
thatformachannel-20Aindiameter.However,thiswould
beseeninelectronmicrographs.Tobesure,othersclaimto
haveseentheexpectedstructure,buttheevidencepresented
(306,307,324)wasnot acceptable(seereference5).Thus,it
seemedunlikelythattheearliermodels(308,322,323)were
correct.
Zampighietal.(327)havereportedrecentlyonagap-
junctionfraction(313)studiedinthin sectionsandbynegative
stainingusingstereo-imageanalysistechniques.Themajor
conclusionreachedwasthattransversechannelscouldnotbe
seenclearlyinsectionsnorinedge-oneviewsofthe junctions
innegative-stainpreparations.Tiltstudiesshowedclearlythat
the-20Apoolsofstrainseeninfrontalviews ofnegative-
stainpreparations did notexistascolumnsrunning through
thetwojunctionalmembranes.
ZampighiandUnwin(328,329)pursuedthesestudiesby
employingaminimal-doseelectronmicroscopetechnique(157,
158).Theyworkedoutthethree-dimensional (3-D) structure
toalevelof18Aresolution.Theyconcluded:
(a)Thejunctionsmayexistintwoforms,AandBdepend-
ingondetergentcontent.
(b)Theconnexonconsistsofsixslightlytwistedprotein
subunits
asymmetricallydisposedacrosseachmem-
branewithconsiderablemassprotrudingfromthebi-
layersurfaceonthe outside butessentiallynoneonthe
cytoplasmicside.TwistisgreaterintheAform.
(c)Thenegativestainwasconcentratedintworegions
betweenthetwojunctionalmembranes.
(d)Abarelydetectable amountofstainpenetratedthe
channelthroughthetwoadjacentmembranes.
(e)Theregulationofsizeofthechannelwasmostprobably
localizedtothecore ofeachbilayer.Theseresultsled
themtopostulateaheuristicmodelforthejunction
with thesixsubunitsarrangeddiagonallyacrossthe
bilayer.Fig.11fromZampighiandUnwin(328)shows
contourmapsofcrosssectionsofonemembraneof
junctionsineach ofthetwostates.
Fromtheaboveitisclearthatthechannelintheclosedstate
isahydrophobicstructureoverall.Itreacts toheavy-metal
stains likeBRinthe purplemembrane.Howcanthisbe
reconciledwith the function ofthe channel?Weknowthatthe
channelmust beabletopasshydrophilicmoleculesupto-20
Aindiameter(294).However,inisolation thechannelsare
hardlypenetrablebymuchsmallerheavy-metalstainmole-
cules.Clearly,thecorepartof thechannelmust beadynamic
structurecapableofmarkedchanges.Theremustbemore
hydrophobicaminoacidresiduesinthecorethanhydrophilic
ones,asinthepurplemembrane,buttheremust besomeway
for thehydrophilicresiduesto bearrangedtomakeatransverse
hydrophilicchannelupto^20Aindiameterintheopen
condition,butstillable toreturntoa verydifferentarrange-
mentin theclosedstate.Itwillbeexcitingtoseehowthe
evidence developsaswelearnmoreaboutthisfascinating
structureandareabletoarriveatapreciseunderstandingof
howit isconstructedandfunctions.
Conclusion
Thischapter hasattemptedtotracethe evolutionofour
ideasaboutthemolecularstructureofcellmembranesas
embodiedinvariousparadigms.TheFMmodelprovideda
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
aldehydefixedepithelial cellsintheileumofanadultrat.After
fixationthejunctionfeaturesacharacteristicbeltlikenetworkof
branchedand anastomosingridges(R)onthePFfaceandcorre-
spondingfurrowsontheEF face.Microvilli(MV)areseenabove.
AnextensivePF faceisseenbelow.x42,380.
asthetransmembranepartofthemolecule.Theyalsonoted
thatthereisa very highmolarratioofcholesterolinthe
junctionsdespitethe detergenttreatment,andsuggestedthatit
playsastructuralrole.
Finbowet al.(319)havereportedanindependentlineof
evidence
whichalsosuggeststhatthemajorgapjunctional
proteinisthe 26,000dcomponent.Thisgrouphadshown(320,
321)thatpartialhepatectomycaused gapjunctionstodisappear
duringapostoperativeperiodof24-28hwithareturnto
normalwithin48h.Theynotedthatthe26,000dcomponent
wasabsentinthe24-28-hperiodandreappearedat48h.There
isnowgeneralagreementthattheprincipalgapjunctional
proteinhasanapparentmolecularweightof26,000d.
Casparetal.(322)andMakowskiet al.(323)conducteda
combinedchemical,electronmicroscope,andX-raydiffraction
study ofisolatedgapjunctionswhichledthemtoproposea
model.Theyreliedheavilyonamicrographpublishedby
Goodenoughin1976(307, 324),interpretedasshowingPTA
fillingan-20Atransversehydrophilicchannelinanisolated
junction.Thisledthemtopostulatetheexistenceofaqueous
-20A-diameter,protein-linedchannelscompletelytraversing
thejunctions.Theysupposedthatflowthroughthesechannels
wasregulatedbyvariationsinthediameterof thechannelin
the regionbetweenthetwomembranes.Theyapplied thename
"connexon"for thecompletechannelsthatrunacrosseach of
thetwomembranesof a given junction.
Ithaslongbeenclearthatsomesortoftransversechannel
structureispresentintheunitmembranesofthe junctions.
Lowenstein,forexample,hasshownthatmoleculesupto20A
indiametercanpassthrough(300).However,electronmicro-
graphsfailed toshowdirectevidenceofanysuch pore.Thus,
ZampighiandRobertson,in1974,(325)foundthatitwas
possibletodegradetheisolatedjunctionselectivelybytreating
itwithEDTAorEGTA.Aftertreatment,the junctionbroke
upintofragmentsthat consistedofonlyafewoftherepeating
units,someofwhichwerefoundtohe ontheirside innegative
stain.Noevidence oftransversechannelswasfound.Achannel
-20Aindiameterwouldbeexpectedtofillandbeseenunder
theseconditions,becausethereisacomparableholeintobacco
mosaicvirusthatfillsreadilywithPTA(326).Itwasconcluded
thatthechannelmustbesmallerthanthe-20Asuggestedby
thesizeofthestainaccumulationbetweenthetwomembranes.
This broughtintofocus theproblemof the nature ofthe
channel.Obviously,onepossibilitywouldbe a tubularstruc-
200s
THEJOURNALOFCELLBIOLOGY"VOLUME91,1981
tureconsisting,perhaps, ofsomethinglikeafl-pleatedsheetof
polypeptidechainsrolledintoacylindricalformwithhydro-
phobicresiduesontheoutsideandhydrophilicresiduesinside
thatformachannel-20Aindiameter.However,thiswould
beseeninelectronmicrographs.Tobesure,othersclaimto
haveseentheexpectedstructure,buttheevidencepresented
(306,307,324)wasnot acceptable(seereference5).Thus,it
seemedunlikelythattheearliermodels(308,322,323)were
correct.
Zampighietal.(327)havereportedrecentlyonagap-
junctionfraction(313)studiedinthin sectionsandbynegative
stainingusingstereo-imageanalysistechniques.Themajor
conclusionreachedwasthattransversechannelscouldnotbe
seenclearlyinsectionsnorinedge-oneviewsofthe junctions
innegative-stainpreparations.Tiltstudiesshowedclearlythat
the-20Apoolsofstrainseeninfrontalviews ofnegative-
stainpreparations did notexistascolumnsrunning through
thetwojunctionalmembranes.
ZampighiandUnwin(328,329)pursuedthesestudiesby
employingaminimal-doseelectronmicroscopetechnique(157,
158).Theyworkedoutthethree-dimensional (3-D) structure
toalevelof18Aresolution.Theyconcluded:
(a)Thejunctionsmayexistintwoforms,AandBdepend-
ingondetergentcontent.
(b)Theconnexonconsistsofsixslightlytwistedprotein
subunits
asymmetricallydisposedacrosseachmem-
branewithconsiderablemassprotrudingfromthebi-
layersurfaceonthe outside butessentiallynoneonthe
cytoplasmicside.TwistisgreaterintheAform.
(c)Thenegativestainwasconcentratedintworegions
betweenthetwojunctionalmembranes.
(d)Abarelydetectable amountofstainpenetratedthe
channelthroughthetwoadjacentmembranes.
(e)Theregulationofsizeofthechannelwasmostprobably
localizedtothecore ofeachbilayer.Theseresultsled
themtopostulateaheuristicmodelforthejunction
with thesixsubunitsarrangeddiagonallyacrossthe
bilayer.Fig.11fromZampighiandUnwin(328)shows
contourmapsofcrosssectionsofonemembraneof
junctionsineach ofthetwostates.
Fromtheaboveitisclearthatthechannelintheclosedstate
isahydrophobicstructureoverall.Itreacts toheavy-metal
stains likeBRinthe purplemembrane.Howcanthisbe
reconciledwith the function ofthe channel?Weknowthatthe
channelmust beabletopasshydrophilicmoleculesupto-20
Aindiameter(294).However,inisolation thechannelsare
hardlypenetrablebymuchsmallerheavy-metalstainmole-
cules.Clearly,thecorepartof thechannelmust beadynamic
structurecapableofmarkedchanges.Theremustbemore
hydrophobicaminoacidresiduesinthecorethanhydrophilic
ones,asinthepurplemembrane,buttheremust besomeway
for thehydrophilicresiduesto bearrangedtomakeatransverse
hydrophilicchannelupto^20Aindiameterintheopen
condition,butstillable toreturntoa verydifferentarrange-
mentin theclosedstate.Itwillbeexcitingtoseehowthe
evidence developsaswelearnmoreaboutthisfascinating
structureandareabletoarriveatapreciseunderstandingof
howit isconstructedandfunctions.
Conclusion
Thischapter hasattemptedtotracethe evolutionofour
ideasaboutthemolecularstructureofcellmembranesas
embodiedinvariousparadigms.TheFMmodelprovideda
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
FIGURE11
Contourmapsoftwoperpendicularsectionsthrough
one
membraneoftwodifferentgapjunctionsshowingtwodifferent
statesinwhichthejunctioncanexist.Arrowheadspointtothe
approximatelocationsofthecytoplasmic(upper)andextracellular
(lower)membranesurfaces.Thezones(C,MandE)refer tothe
cytoplasm,membrane, andextracellularspace,respectively .Con-
tourscorrespondingtostainexcludingregions(negativecontours)
aredrawnascontinuouslines.Contourscorrespondingtostain-
filledregions(0andpositivecontours)aredrawnasbrokenlines.
Thesectionscontaintwounitcells inthehorizontaldirectionand
halfofthejunctionintheverticaldirection.Notetheconcentration
ofstaininthecentralregionalongtheconnectionaxismainly
betweenthetwomembranes.Asmallerconcentrationofstain
occursatthe peripheryofeachsubunit.Twostatesareshown.
Thereisaslightopeninginthecenteroftheconnexontothe top
state(a).Inthetransitionbetweenthetwostatesmattermoves
towardtheconnexonaxis (verticalarrow)toclosetheslightopening
inthecytoplasmicsurfaceingoingfroma tob .
veryusefulmembraneparadigmfor the1970s.Itfocused
attentionontransmembraneproteinsandmembranefluidity
at
a timewhenthesefeaturesofmembraneswerecomingto
the
forefrontofmembraneresearch.Themodelservedavery
usefulpurposeinemphasizingtheimportanceoftransmem-
braneproteins.However,aswithallmodels,oversimplifica-
tionsinevitablyoccurred.Itbecameobviousalmostimmedi-
atelythatthefluidityelementhadbeengreatlyexaggerated
andthatdepictingmuchof thebilayerasnakedwasincorrect.
Themodelalsohasbeenmisleadinginsuggestingthatthereis
quantitativelymoreproteininthehydrophobiccore of the
bilayerthaninmostmembranes.Also,themodelhasledsome
toregardtheubiquitousIMPSasmetal-plated proteinmole-
cules.ItisnowapparentthatIMPsaremorecomplex.Even
whenrelatedtotransmembraneproteins,theydonotgivea
faithfulrepresentationofthoseproteins.Polypeptidesarepres-
entinthebilayercoreinmuchsmaller quantitythanthe
numberandsizeoftheIMPssuggest.Finally,theFMmodel
hassupportedtheconceptofpermanenthydrophilic,water-
filled,transmembranechannels(12,316),forwhichthereisno
structuralevidence,It isthusquite clearthattheFMmodel
needstoberevisedinsignificantways,althoughsomeofits
featuresremainvalid.Fig.12presentsina highlyschematic
FIGURE12
Highlyschematicdiagramofmodelofacellmembrane.
Thelipidbilayercoreisrepresentedbythe joinedcircleandrectan-
gularfigures.Theasymmetryinthelipidbilayerdiscussedinthe
textisrepresentedbyfillinginthenonpolar carbonchainregions
(rectangles)andheadgroups(circles)ofthelipidmoleculesinone
halfofthebilayer .Theprotein constituentsarecross-hatcheddif-
ferentlyto indicatetheasymmetryoftheinnerandouterprotein
componentsandthe existenceoftransmembraneproteincompo-
nentsisindicatedby athirdcrosshatchpattern.Thepresenceof
sugarresiduesintheexternalsurfaceofthemembraneisrepre-
sentedbythebranchedchainsofjoinedhexagonsintheexternal
surface.Themoleculesaredrawnapproximatelytoscalebutvery
schematically.Thebilayerisabout50Athickand eachprotein
monolayerisabout20Athick.Noeffortismadetoshowdifferent
kindsoflipidmoleculesbut thefactthatthelipidmoleculesarein
arelativeliquidstateisindicatedbyshowingdifferentprojections
asseenindifferent statesofrotation.Fluidityowingtoflexingof
thehydrocarbon chainsisnotshown.
fashion amodel,referredto asthehydrophobicbarriermodel,
incorporatingfeaturesofallcontemporarymembranemodels.
ACKNOWLEDGMENTS
ThemicrographinFigure4 waskindlysuppliedbyDr.H.PingBeall.
REFERENCES
1.Davis,B.D.,andP-C.Tai.1980.Nature(Loud.).283:433-438.
2.Stoeckenius,W.,R.H.Lozier,andR.A.Bogomolni.1979.Biochim.
Biophys.Acta.505:215-278.
3.Lodish,H.F.,andJ.E.Rothman.1979.Sci.Am.240:48-63.
4.Robertson,J.D.1978.InPhysiologyofMembraneDisorders.T.Andreoli,
J.Hoffman,andD.Fanestil,editors.PlenumPress,NewYork.1-25.
5.Robertson,J.D .1978.InPhysiologyofMembraneDisorders.T.Andreoli,
J .Hoffman,andD.Fanestil,editors.PlenumPress,NewYork.61-93.
6.Guidotti,G.1978.InPhysiologyofMembraneDisorders.T.J.Andreoli,J.
F.Hoffman, andD.D.Fanestil,editors.PlenumPress.NewYork.49-60.
7 .Rothman,J.E.,andJ.Lenard .1977.Science(Wash.D.C.).195:743-753.
8.Abrahamsson,S.,and1.Pascher,editors.1977.InStructureofBiological
Membranes.PlenumPress,NewYork.580pp.
9.Robertson,J.D.1975.InTheNervousSystem,Vol.I.BasicNeurosciences.
D.B.Tower,editor.RavenPress,NewYork, 43-58.
10.Steck,T.L.1974.J.CellBiol.62:1-19.
11.Singer,S.J.1971.InStructureandFunctionofBiologicalMembranes.L .
1.Rothfield,editor.AcademicPress,Inc.,NewYork.145-222.
12.Singer,S.J.1974.Annu.Rev.Biochem.43:805-833.
13 .Hendler,R.W.1971.Physiol.Rev.51:66-97.
14 .Stoeckenius,W.,andD.M.Engelman.1969.J.CellBiol.42:613-646.
15 .Neumann,E.,andJ.Bernhardt.1977.Annu.Rev.Biochem.46:117-141.
16.Ehrenstein,G.,andH.Lecar.1977.Q.Rev.Biophys.10:1-34.
17.Tada,M.,T.Yamamoto,andY .Tonomura.1978.Physiol.Rev.58:1-79.
18.Sachs,G.,J.G.Spenuey,andM.Lewin.1978.Physiol.Rev.58:106-173.
ROBERTSON
MembraneStructure
2015
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
Contourmapsoftwoperpendicularsectionsthrough
one
membraneoftwodifferentgapjunctionsshowingtwodifferent
statesinwhichthejunctioncanexist.Arrowheadspointtothe
approximatelocationsofthecytoplasmic(upper)andextracellular
(lower)membranesurfaces.Thezones(C,MandE)refer tothe
cytoplasm,membrane, andextracellularspace,respectively .Con-
tourscorrespondingtostainexcludingregions(negativecontours)
aredrawnascontinuouslines.Contourscorrespondingtostain-
filledregions(0andpositivecontours)aredrawnasbrokenlines.
Thesectionscontaintwounitcells inthehorizontaldirectionand
halfofthejunctionintheverticaldirection.Notetheconcentration
ofstaininthecentralregionalongtheconnectionaxismainly
betweenthetwomembranes.Asmallerconcentrationofstain
occursatthe peripheryofeachsubunit.Twostatesareshown.
Thereisaslightopeninginthecenteroftheconnexontothe top
state(a).Inthetransitionbetweenthetwostatesmattermoves
towardtheconnexonaxis (verticalarrow)toclosetheslightopening
inthecytoplasmicsurfaceingoingfroma tob .
veryusefulmembraneparadigmfor the1970s.Itfocused
attentionontransmembraneproteinsandmembranefluidity
at
a timewhenthesefeaturesofmembraneswerecomingto
the
forefrontofmembraneresearch.Themodelservedavery
usefulpurposeinemphasizingtheimportanceoftransmem-
braneproteins.However,aswithallmodels,oversimplifica-
tionsinevitablyoccurred.Itbecameobviousalmostimmedi-
atelythatthefluidityelementhadbeengreatlyexaggerated
andthatdepictingmuchof thebilayerasnakedwasincorrect.
Themodelalsohasbeenmisleadinginsuggestingthatthereis
quantitativelymoreproteininthehydrophobiccore of the
bilayerthaninmostmembranes.Also,themodelhasledsome
toregardtheubiquitousIMPSasmetal-plated proteinmole-
cules.ItisnowapparentthatIMPsaremorecomplex.Even
whenrelatedtotransmembraneproteins,theydonotgivea
faithfulrepresentationofthoseproteins.Polypeptidesarepres-
entinthebilayercoreinmuchsmaller quantitythanthe
numberandsizeoftheIMPssuggest.Finally,theFMmodel
hassupportedtheconceptofpermanenthydrophilic,water-
filled,transmembranechannels(12,316),forwhichthereisno
structuralevidence,It isthusquite clearthattheFMmodel
needstoberevisedinsignificantways,althoughsomeofits
featuresremainvalid.Fig.12presentsina highlyschematic
FIGURE12
Highlyschematicdiagramofmodelofacellmembrane.
Thelipidbilayercoreisrepresentedbythe joinedcircleandrectan-
gularfigures.Theasymmetryinthelipidbilayerdiscussedinthe
textisrepresentedbyfillinginthenonpolar carbonchainregions
(rectangles)andheadgroups(circles)ofthelipidmoleculesinone
halfofthebilayer .Theprotein constituentsarecross-hatcheddif-
ferentlyto indicatetheasymmetryoftheinnerandouterprotein
componentsandthe existenceoftransmembraneproteincompo-
nentsisindicatedby athirdcrosshatchpattern.Thepresenceof
sugarresiduesintheexternalsurfaceofthemembraneisrepre-
sentedbythebranchedchainsofjoinedhexagonsintheexternal
surface.Themoleculesaredrawnapproximatelytoscalebutvery
schematically.Thebilayerisabout50Athickand eachprotein
monolayerisabout20Athick.Noeffortismadetoshowdifferent
kindsoflipidmoleculesbut thefactthatthelipidmoleculesarein
arelativeliquidstateisindicatedbyshowingdifferentprojections
asseenindifferent statesofrotation.Fluidityowingtoflexingof
thehydrocarbon chainsisnotshown.
fashion amodel,referredto asthehydrophobicbarriermodel,
incorporatingfeaturesofallcontemporarymembranemodels.
ACKNOWLEDGMENTS
ThemicrographinFigure4 waskindlysuppliedbyDr.H.PingBeall.
REFERENCES
1.Davis,B.D.,andP-C.Tai.1980.Nature(Loud.).283:433-438.
2.Stoeckenius,W.,R.H.Lozier,andR.A.Bogomolni.1979.Biochim.
Biophys.Acta.505:215-278.
3.Lodish,H.F.,andJ.E.Rothman.1979.Sci.Am.240:48-63.
4.Robertson,J.D.1978.InPhysiologyofMembraneDisorders.T.Andreoli,
J.Hoffman,andD.Fanestil,editors.PlenumPress,NewYork.1-25.
5.Robertson,J.D .1978.InPhysiologyofMembraneDisorders.T.Andreoli,
J .Hoffman,andD.Fanestil,editors.PlenumPress,NewYork.61-93.
6.Guidotti,G.1978.InPhysiologyofMembraneDisorders.T.J.Andreoli,J.
F.Hoffman, andD.D.Fanestil,editors.PlenumPress.NewYork.49-60.
7 .Rothman,J.E.,andJ.Lenard .1977.Science(Wash.D.C.).195:743-753.
8.Abrahamsson,S.,and1.Pascher,editors.1977.InStructureofBiological
Membranes.PlenumPress,NewYork.580pp.
9.Robertson,J.D.1975.InTheNervousSystem,Vol.I.BasicNeurosciences.
D.B.Tower,editor.RavenPress,NewYork, 43-58.
10.Steck,T.L.1974.J.CellBiol.62:1-19.
11.Singer,S.J.1971.InStructureandFunctionofBiologicalMembranes.L .
1.Rothfield,editor.AcademicPress,Inc.,NewYork.145-222.
12.Singer,S.J.1974.Annu.Rev.Biochem.43:805-833.
13 .Hendler,R.W.1971.Physiol.Rev.51:66-97.
14 .Stoeckenius,W.,andD.M.Engelman.1969.J.CellBiol.42:613-646.
15 .Neumann,E.,andJ.Bernhardt.1977.Annu.Rev.Biochem.46:117-141.
16.Ehrenstein,G.,andH.Lecar.1977.Q.Rev.Biophys.10:1-34.
17.Tada,M.,T.Yamamoto,andY .Tonomura.1978.Physiol.Rev.58:1-79.
18.Sachs,G.,J.G.Spenuey,andM.Lewin.1978.Physiol.Rev.58:106-173.
ROBERTSON
MembraneStructure
2015
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
19.Finn,A.L.1976.Physiol.Rev.56:453-464.
20.Brenner,B.M.,C.Baylis,andW.M.Deen.1976.Physiol Rev.56:502-534.
21.Narahashi,T.1974.PhysiolRev.54:813-889.
22.Mueller,P.,D.V.Rudin,M.Glen,andW.C.Wescott.1962.Circulation.
26:1167-1171.
23.Mueller,P.,D.O.Rudin,H.T .Tien,andW.C.Wescott.1963.J.Phys.
Chem.67:534-535.
24.Mueller,P.,andD.O.Rudin.1967.Nature(Land).213:603-604.
25.Thompson,T.E.1967.Protoplasma.63:194-196.
26.Haydon,D.A.,andS.B.Hladky.1972.Q.Rev.Biophys.5:187-282.
27.Montal,M.,andP.Mueller .1972.Proc.Nail.Acad.Sci.U.S.A .69 :3561-
3566.
28.Latorre,R.,andO.Alvarez.1980.Annu.Rev.Biochem.Inpress.
29.Hall,J.E.1978 .InMembraneTransportinBiology.G.Giebisch,D.C.
Tosteson,andH.H.Ussing,editors.Springer-Verlag,Berlin.475-531.
30.Tien,H.T.1974 .InBilayerLipidMembranes(BLM),TheoryandPractice.
Marcel Dekker,Inc.,NewYork.655 pp.
31.Korenbrot,J.I.1977.Annu.Rev.Physiol.39:19-49.
32.Semenza,G.,andE.Carafoli,editors .1977.BiochemistryofMembrane
Transport.Fed.Eur.Biochem.Soc.Symp.(Bert).No.42.669 pp.
33.Wickner,W.1979.Annu.Rev.Biochem.48:23-45.
34.Heidmann,T.,andJ-P.Changeaux.1978.Annu.Rev.Biochem.47:317-357.
35.Marchesi,V.T.,andH.Furthmeyer.1976.Annu.Rev.Biochem.45:667-
698.
36.Wilson,E .B.1925.InTheCellinDevelopmentandHeredity,3rded.,
MacmillanCo.,NewYork.1-20.
37.Bowman,W.1840.Philos.Trans.RSoc.Land.130:457-501.
38.Overton,E.1895.Vierteljahrschr.Naturforsch.Ges.Zur.40:159-201.
39.Overton,E.1899.Viertejahrschr.Naturforseh.Ges.Zur.44:669-701.
40.Bernstein,J.1902.Arch.GesamtePhysiol.Mens.Tiere(Piers)92:521-
562.
41.Bernstein,J.1868.Arch.GesamtePhysiol.Mens.Tiere(Pflugers)1:173-207.
42.Hober,R.1910.Arch.GesamtePhysiol.Mens.Tiere (Pflugers)133:237-259.
43.H6ber,R.1912.Arch.GesamtePhysiol.Mens.Tiere(P(hugers)148:189-221.
44.Hober,R.1913.Arch.GesamtePhysiol.Mens.Tiere(Pflugers)150:15-45.
45.Fricke,H.1923.Phys.Rev.21:708-709.
46.Fricke,H.1925.J.Gen.Physiol9:137.
47.Rayleigh,Lord.1890.Proc.RSac.47:364-366.
48.Devaux,H.1913.Transl .fromRev.Gen.Sci.Pures Appl.Annu .Rep.
SmithsonianInst.24(4):261-273.
49.Langmuir,I.1917.J.Am.Chem.Soc.39:1848-1906.
50.Harkins,W.D.1917.J.Am.ChemSoc.39:354 .
51.Harkins,W.D.1917.J.Am.Chem.Soc.39:541.
52.Tanford,C.1979.InTheHydrophobicEffect :FormationofMicellesand
BiologicalMembranes,2ndedition.JohnWileyandSons,NewYork.200
pp-
53.Gorter,E.,andR.Grendel.1925.J.Exp.Med.41:439-443.
54.Schmitt,F.O.,R.S .Bear,andE.Ponder.1837.J.Cell.Comp.Physiol9 :
89-92.
55.Schmitt,F.O.,R.S.Bear,andE.Ponder.1938.J.Cell.Comp.Physiol.11:
309-313.
56.Chambers,R.,andM.J.Kopac.1937.J.Cell.Comp.Physiol.9:331-345.
57.Kopac,M.J.,andR.Chambers.1937.J.Cell.Comp.Physiol.9:345-361.
58 .Davson, J.,andJ.F.Danielh.1943.InThePermeabilityofNatural
Membranes.CambridgeUniversityPress,Cambridge.63.
59.Mudd,S.,andE.B.H.Mudd.1931.J.Exp.Med.53:733-750.
60.Robertson,J .D.1975.InTheNervousSystem,Vol.I.BasicNeurosciences.
D.B.Tower,editor.RavenPress,NewYork.43-58.
61.Cole,K.S.1932.J.Cell.Comp.Physiol1:1-9.
62.Harvey,E.M.,andH.Shapiro.1934.J.CeltComp.Physiol. 5:255-267.
63.Danielh,J.F.,andE.N.Harvey.1934 .JCellComp.Physiol5:483-494.
64.Langmuir, I.,andD .F.Waugh.1938.J.Gen.Physiol.21:745-755.
65.Devaux,H.1935.C.RAcad.Sci.201:109-Ill.
66.Danielli,J.F.1938.ColdSpringHarborSymp.Quant.Biol.6:190-195.
67.Kopac,M.J.1938.Biol.Bull.(WoodsHole).75:351.
68.Kopac,M.J.1950.Ann.N.Y.Acad.Sci.50:870-909.
69.Trumit,H.J.1960.J.ColloidSci.15:1-13.
70.VanDeenen,L .L.M.,U.M.T.Houtsmuller,G .H.de Haas,andE.
Muller .1962.J.Pharm.Pharmacol14:429.
71.Turner,S.R.,M.Litt,andW.S.Lynn.1973.J.ColloidInterfaceSci.48:
100-104 .
72.Cole,K.S.,andH.J.Curtis.1950 .Med.Phys.II:82-90.
73.Cole,K.S.1935.J.Gen.Physiol.18:877.
74.Iida,T.T.1943a.J.Fac.Sci.Imp.Univ.TokyoSect.IV.Zool 6:141-151.
75.Iida,T.T.1943a.J.Fac.Sci.Imp.Univ.TokyoSect.IV.Zool.6:114-115.
76.Katz,B.1949 .Arch.Sci.Physiol.3:449-460.
77.Rothschild,Lord.1957.J.Biophys.Biochem.Cytol.3:103-110.
78.Robertson,J.D.1956.J.Biophys.Biochem.Cytol2:369-379.
79.Porter,K.R.,andG.E.Palade.1957.J.Biophys.Biochem.Cytol.3:269-
300.
80.Peachey,L.D .1965.J.CellBiol.25:209-231.
81.Huxley,A.F.1971.Proc.RSoc.Land.B.Biol.Sci.178:1-27.
82.Danielli,J.F.,andH.Davson .1935.J.Cell.Comp.Physiol.5:495-508.
83.Danielli,J.F.1954.ColstonPap.7:1-14.
84.Danielli,J.F.1935.J.Gen.Physiol.19:19-22.
85.Danielli,J.F.1936.J.Cell.Comp.Physiol. 7:393-408.
2025
THEJOURNALOFCELLBIOLOGY"VOLUME91,1981
86.Waugh,D.F.,andF.O.Schmitt.1940.ColdSpringHarborSymp.Quant.
Biol8:233-241.
87.Luft,J.H.1956.J.Biophys.Biochem.Cytol2:799-801.
88.Glauert,A .M.,G.E.Rogers,andR.H.Glauert.1956.Nature (Loud.).178:
803.
89.Robertson,J.D.1957.J.Physiol(Land.).140:58-59.
90.Robertson,J.D.1957.J.Biophys.Biochem.Cytol 3:1043-1048.
91.Robertson,J.D.1958.J.Biophys.Biochem.Cytol4:39-46.
92.Robertson,J.D .1958.J.Biophys.Biochem.Cytol4:349-364.
93.Robertson,J.D.1959.Biochem.Soc.Symp.16:3-43.
94.Robertson,J.D.1960.Prog.Biophys.Biophys.Chem11:343-418 .
95.Robertson,J.D.1960.Int.Conf.ElectronMicrose.Proc.4th159-171.
96.Robertson,J.D.1960 .InMolecularBiology.D.Nachmansohn,editor.
AcademicPress,Inc.,NewYork.87-151 .
97.Robertson,J.D.1963.In CellularMembranesinDevelopment:Proceedings
oftheXXIISymposiumoftheSocietyfor theStudy ofDevelopmentand
Growth.MichaelLocke,editor.Academicpress,Inc.,NewYork.1-81.
98.Heckman,C.W.,andJ.H.Barmett.1973.J.Ultrastruct.Res.42:156-179.
99.vanCriegee,R.1936.Ann.Chem.(JustusLiebigs).522:75-96.
100.Stoeckenius,W.1962.J.Cell Biol.12:221-229.
101.Luzzati,V.,andF.Husson.1962 .J.Biophys.Biochem.Cytol.12:207-219.
102.Geren,B.B.1954.Exp.CellRes.7:588.
103.Robertson,J.D.1955.J.Biophys.Biochem.Cytol.2:369.
104.Schmitt,F.O.,R.S.Bear,andK.H.Palmer.1941 .J.Cell.Comp.Physiol.
18:39.
105.Finean,J.B.1956.InProceedings2ndInternationalConference(Ghent).
G.PopjakandE.LeBreton,editors.ButterworthsScientificPub.,London.
127-131.
106.Schmidt,W.J.1937.Z.Wiss.Mikrosk.Tech.54:159.
107.Moody,M.1963.Science(Wash.D.C.)142:1173.
108.Caspar,D.L.D.,andD.A.Kirschner.1971.Nature(Land.).231:46-52.
109.Revel,J.P.,L.Napolitano,andD.W.Fawcett.1960.JBiophys.Biochem.
Cytol8:575-589.
110.Robertson,J.D.1962.Sci.Am.206:64-72.
111.Kom,E.D.1966.Science(Wash .D.C.).153:1491-1498.
112.Sjostrand,F.S.1965 .J.Ulstrastruct.Res.9:340-361.
113.Sjdstrand,F.S.1969 .In StructuralandFunctionalAspectsofLipoprotein
in LivingSystems.E.TriaandA.M.Scanueditors.AcademicPress,Inc.
NewYork.73-139.
114.Lenard,J.,and S.J.Singer.1966.Proc.Nail.Acad.Sci.U.S.A.56 :1828-
1835.
115.Green,D.E.,andT.Oda.1961.J.Biochem.(Tokyo).49:742-757.
116.Green,D.E.,andJ.F.Pardue.1966.Biochemistry.55:1295-1302.
117.Benson,A.A.1964.Annu.Rev.PlantPhysiol.15:1-16.
118.Benson,A.A.1966.J.Am.OilChem.Soc.43:265-270.
119.Sjdstrand,F.S.1976 .J.Ultrastruct.Res.55:271-280.
120.Sjostrand,F.S.,andL.Barajas.1968.J.Ultrastruct.Res.25:121.
121.Sjostrand,F.S.,andL.Barajas.1970.J.Ultrastruet.Res.32:293-306.
122.Ohad,I.,P.Siekevitz,andG.E.Palade.1967.J.CellBiol.35:521-552.
123 .Ohad,I.,P.Siekevitz,andG.E.Palade.1967.J.CellBiol.35:553-584.
124.Dallner,G.,P.Siekevitz,andG.E.Palade.1966.J.CellBiol.30:73-96;97-
117.
125.Omura,T.,P.Siekevitz,andG.E.Palade.1967.J.Biol.Chem.242:2389-
2396.
126.Hoober,K.,P.Siekevitz,andG.E .Palade .1969.J.Biol.Chem.244:2621-
2631.
127 .Schor,S.,P.Siekevitz,andG.E.Palade.1970.Proc.Nail.Acad.Sci.U.S.
A.66:174-179.
128.Back,K.W.,P.Siekevitz,andG.E.Palade.1971 .J.Biol.Chem.246:188-
195.
129.Steere,R.L.1957.J.Biophys.Biochem.Cytol.3:45-60.
130.Moor,H.,K.Muhlethaler,H .Waldner,andA.Frey-Wysshng.1961.J.Cell
Biol.10:1-15.
131.Branton,D.1966.Proc.Nat.Acad.Sci.U.S.A.55:1048-1056.
132 .Branton,D.,andR.B.Park.1967.J.Ultrastruct.Res.19 :283-303.
133.Branton,D.1971.Philos.Trans:RSoc.Land.B.Biol.Sci.261:193-138.
134.daSilva,P.,andBranton,D .1970.J.Cell Biol.45:598-605.
135.Branton,D.1967.Exp.CellRes.45:703-707.
136.Branton,D.,andD.W.Deamer.1972.Protoplasmatologia.2(Sect.1):1-70.
137.Deamer,D.,andD.Branton .1967.Science(Wash.D.C).158:655-656.
138.Deamer,D.,R.Leonard,A.Tardieu,andD.Branton.1970 .Biochem.
Biophys.Acta.219:47-60.
139 .Mateu,L.,V.Luzzati,Y.London,R.M.Gould,andF.G.A.Vasseberg.
1973.J.Mal.Biol.75:697-709.
140.Gulik-Krzywieki,T.1975.Biochem.Biophys.Acta.415:1-28.
141.Clark,A.W.,andD.Branton.1968.Fracturefacesinfrozenoutersegments
fromtheguineapigretina.ZZellforsch.Mikrosk.Anal.91:586-603.
142.Leeson,T .S.1970.Ratretinalrods:freeze-fracture replicasofouter
segments.Can.J.Ophthalmol.5:91-107.
143.Leeson,T.S.1971.J.Anat.108:147-157.
144.Corless,J.M.,W.H.CobbsIII,M.J.Costello,andJ.D.Robertson.1976.
Exp.Eye Res.23:295-324.
145.Corless,J.M.,andM.J.Costello,1980.Biophys.J.25:289A.
146.Singer,S.J.,andG.L.Nicholson .1972.Science(Wash .D.C.).175:720-
731.
147.Frye,L,D.,andM.Edidin.1970.J.CellSci.7:319-335.
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
20.Brenner,B.M.,C.Baylis,andW.M.Deen.1976.Physiol Rev.56:502-534.
21.Narahashi,T.1974.PhysiolRev.54:813-889.
22.Mueller,P.,D.V.Rudin,M.Glen,andW.C.Wescott.1962.Circulation.
26:1167-1171.
23.Mueller,P.,D.O.Rudin,H.T .Tien,andW.C.Wescott.1963.J.Phys.
Chem.67:534-535.
24.Mueller,P.,andD.O.Rudin.1967.Nature(Land).213:603-604.
25.Thompson,T.E.1967.Protoplasma.63:194-196.
26.Haydon,D.A.,andS.B.Hladky.1972.Q.Rev.Biophys.5:187-282.
27.Montal,M.,andP.Mueller .1972.Proc.Nail.Acad.Sci.U.S.A .69 :3561-
3566.
28.Latorre,R.,andO.Alvarez.1980.Annu.Rev.Biochem.Inpress.
29.Hall,J.E.1978 .InMembraneTransportinBiology.G.Giebisch,D.C.
Tosteson,andH.H.Ussing,editors.Springer-Verlag,Berlin.475-531.
30.Tien,H.T.1974 .InBilayerLipidMembranes(BLM),TheoryandPractice.
Marcel Dekker,Inc.,NewYork.655 pp.
31.Korenbrot,J.I.1977.Annu.Rev.Physiol.39:19-49.
32.Semenza,G.,andE.Carafoli,editors .1977.BiochemistryofMembrane
Transport.Fed.Eur.Biochem.Soc.Symp.(Bert).No.42.669 pp.
33.Wickner,W.1979.Annu.Rev.Biochem.48:23-45.
34.Heidmann,T.,andJ-P.Changeaux.1978.Annu.Rev.Biochem.47:317-357.
35.Marchesi,V.T.,andH.Furthmeyer.1976.Annu.Rev.Biochem.45:667-
698.
36.Wilson,E .B.1925.InTheCellinDevelopmentandHeredity,3rded.,
MacmillanCo.,NewYork.1-20.
37.Bowman,W.1840.Philos.Trans.RSoc.Land.130:457-501.
38.Overton,E.1895.Vierteljahrschr.Naturforsch.Ges.Zur.40:159-201.
39.Overton,E.1899.Viertejahrschr.Naturforseh.Ges.Zur.44:669-701.
40.Bernstein,J.1902.Arch.GesamtePhysiol.Mens.Tiere(Piers)92:521-
562.
41.Bernstein,J.1868.Arch.GesamtePhysiol.Mens.Tiere(Pflugers)1:173-207.
42.Hober,R.1910.Arch.GesamtePhysiol.Mens.Tiere (Pflugers)133:237-259.
43.H6ber,R.1912.Arch.GesamtePhysiol.Mens.Tiere(P(hugers)148:189-221.
44.Hober,R.1913.Arch.GesamtePhysiol.Mens.Tiere(Pflugers)150:15-45.
45.Fricke,H.1923.Phys.Rev.21:708-709.
46.Fricke,H.1925.J.Gen.Physiol9:137.
47.Rayleigh,Lord.1890.Proc.RSac.47:364-366.
48.Devaux,H.1913.Transl .fromRev.Gen.Sci.Pures Appl.Annu .Rep.
SmithsonianInst.24(4):261-273.
49.Langmuir,I.1917.J.Am.Chem.Soc.39:1848-1906.
50.Harkins,W.D.1917.J.Am.ChemSoc.39:354 .
51.Harkins,W.D.1917.J.Am.Chem.Soc.39:541.
52.Tanford,C.1979.InTheHydrophobicEffect :FormationofMicellesand
BiologicalMembranes,2ndedition.JohnWileyandSons,NewYork.200
pp-
53.Gorter,E.,andR.Grendel.1925.J.Exp.Med.41:439-443.
54.Schmitt,F.O.,R.S .Bear,andE.Ponder.1837.J.Cell.Comp.Physiol9 :
89-92.
55.Schmitt,F.O.,R.S.Bear,andE.Ponder.1938.J.Cell.Comp.Physiol.11:
309-313.
56.Chambers,R.,andM.J.Kopac.1937.J.Cell.Comp.Physiol.9:331-345.
57.Kopac,M.J.,andR.Chambers.1937.J.Cell.Comp.Physiol.9:345-361.
58 .Davson, J.,andJ.F.Danielh.1943.InThePermeabilityofNatural
Membranes.CambridgeUniversityPress,Cambridge.63.
59.Mudd,S.,andE.B.H.Mudd.1931.J.Exp.Med.53:733-750.
60.Robertson,J .D.1975.InTheNervousSystem,Vol.I.BasicNeurosciences.
D.B.Tower,editor.RavenPress,NewYork.43-58.
61.Cole,K.S.1932.J.Cell.Comp.Physiol1:1-9.
62.Harvey,E.M.,andH.Shapiro.1934.J.CeltComp.Physiol. 5:255-267.
63.Danielh,J.F.,andE.N.Harvey.1934 .JCellComp.Physiol5:483-494.
64.Langmuir, I.,andD .F.Waugh.1938.J.Gen.Physiol.21:745-755.
65.Devaux,H.1935.C.RAcad.Sci.201:109-Ill.
66.Danielli,J.F.1938.ColdSpringHarborSymp.Quant.Biol.6:190-195.
67.Kopac,M.J.1938.Biol.Bull.(WoodsHole).75:351.
68.Kopac,M.J.1950.Ann.N.Y.Acad.Sci.50:870-909.
69.Trumit,H.J.1960.J.ColloidSci.15:1-13.
70.VanDeenen,L .L.M.,U.M.T.Houtsmuller,G .H.de Haas,andE.
Muller .1962.J.Pharm.Pharmacol14:429.
71.Turner,S.R.,M.Litt,andW.S.Lynn.1973.J.ColloidInterfaceSci.48:
100-104 .
72.Cole,K.S.,andH.J.Curtis.1950 .Med.Phys.II:82-90.
73.Cole,K.S.1935.J.Gen.Physiol.18:877.
74.Iida,T.T.1943a.J.Fac.Sci.Imp.Univ.TokyoSect.IV.Zool 6:141-151.
75.Iida,T.T.1943a.J.Fac.Sci.Imp.Univ.TokyoSect.IV.Zool.6:114-115.
76.Katz,B.1949 .Arch.Sci.Physiol.3:449-460.
77.Rothschild,Lord.1957.J.Biophys.Biochem.Cytol.3:103-110.
78.Robertson,J.D.1956.J.Biophys.Biochem.Cytol2:369-379.
79.Porter,K.R.,andG.E.Palade.1957.J.Biophys.Biochem.Cytol.3:269-
300.
80.Peachey,L.D .1965.J.CellBiol.25:209-231.
81.Huxley,A.F.1971.Proc.RSoc.Land.B.Biol.Sci.178:1-27.
82.Danielli,J.F.,andH.Davson .1935.J.Cell.Comp.Physiol.5:495-508.
83.Danielli,J.F.1954.ColstonPap.7:1-14.
84.Danielli,J.F.1935.J.Gen.Physiol.19:19-22.
85.Danielli,J.F.1936.J.Cell.Comp.Physiol. 7:393-408.
2025
THEJOURNALOFCELLBIOLOGY"VOLUME91,1981
86.Waugh,D.F.,andF.O.Schmitt.1940.ColdSpringHarborSymp.Quant.
Biol8:233-241.
87.Luft,J.H.1956.J.Biophys.Biochem.Cytol2:799-801.
88.Glauert,A .M.,G.E.Rogers,andR.H.Glauert.1956.Nature (Loud.).178:
803.
89.Robertson,J.D.1957.J.Physiol(Land.).140:58-59.
90.Robertson,J.D.1957.J.Biophys.Biochem.Cytol 3:1043-1048.
91.Robertson,J.D.1958.J.Biophys.Biochem.Cytol4:39-46.
92.Robertson,J.D .1958.J.Biophys.Biochem.Cytol4:349-364.
93.Robertson,J.D.1959.Biochem.Soc.Symp.16:3-43.
94.Robertson,J.D.1960.Prog.Biophys.Biophys.Chem11:343-418 .
95.Robertson,J.D.1960.Int.Conf.ElectronMicrose.Proc.4th159-171.
96.Robertson,J.D.1960 .InMolecularBiology.D.Nachmansohn,editor.
AcademicPress,Inc.,NewYork.87-151 .
97.Robertson,J.D.1963.In CellularMembranesinDevelopment:Proceedings
oftheXXIISymposiumoftheSocietyfor theStudy ofDevelopmentand
Growth.MichaelLocke,editor.Academicpress,Inc.,NewYork.1-81.
98.Heckman,C.W.,andJ.H.Barmett.1973.J.Ultrastruct.Res.42:156-179.
99.vanCriegee,R.1936.Ann.Chem.(JustusLiebigs).522:75-96.
100.Stoeckenius,W.1962.J.Cell Biol.12:221-229.
101.Luzzati,V.,andF.Husson.1962 .J.Biophys.Biochem.Cytol.12:207-219.
102.Geren,B.B.1954.Exp.CellRes.7:588.
103.Robertson,J.D.1955.J.Biophys.Biochem.Cytol.2:369.
104.Schmitt,F.O.,R.S.Bear,andK.H.Palmer.1941 .J.Cell.Comp.Physiol.
18:39.
105.Finean,J.B.1956.InProceedings2ndInternationalConference(Ghent).
G.PopjakandE.LeBreton,editors.ButterworthsScientificPub.,London.
127-131.
106.Schmidt,W.J.1937.Z.Wiss.Mikrosk.Tech.54:159.
107.Moody,M.1963.Science(Wash.D.C.)142:1173.
108.Caspar,D.L.D.,andD.A.Kirschner.1971.Nature(Land.).231:46-52.
109.Revel,J.P.,L.Napolitano,andD.W.Fawcett.1960.JBiophys.Biochem.
Cytol8:575-589.
110.Robertson,J.D.1962.Sci.Am.206:64-72.
111.Kom,E.D.1966.Science(Wash .D.C.).153:1491-1498.
112.Sjostrand,F.S.1965 .J.Ulstrastruct.Res.9:340-361.
113.Sjdstrand,F.S.1969 .In StructuralandFunctionalAspectsofLipoprotein
in LivingSystems.E.TriaandA.M.Scanueditors.AcademicPress,Inc.
NewYork.73-139.
114.Lenard,J.,and S.J.Singer.1966.Proc.Nail.Acad.Sci.U.S.A.56 :1828-
1835.
115.Green,D.E.,andT.Oda.1961.J.Biochem.(Tokyo).49:742-757.
116.Green,D.E.,andJ.F.Pardue.1966.Biochemistry.55:1295-1302.
117.Benson,A.A.1964.Annu.Rev.PlantPhysiol.15:1-16.
118.Benson,A.A.1966.J.Am.OilChem.Soc.43:265-270.
119.Sjdstrand,F.S.1976 .J.Ultrastruct.Res.55:271-280.
120.Sjostrand,F.S.,andL.Barajas.1968.J.Ultrastruct.Res.25:121.
121.Sjostrand,F.S.,andL.Barajas.1970.J.Ultrastruet.Res.32:293-306.
122.Ohad,I.,P.Siekevitz,andG.E.Palade.1967.J.CellBiol.35:521-552.
123 .Ohad,I.,P.Siekevitz,andG.E.Palade.1967.J.CellBiol.35:553-584.
124.Dallner,G.,P.Siekevitz,andG.E.Palade.1966.J.CellBiol.30:73-96;97-
117.
125.Omura,T.,P.Siekevitz,andG.E.Palade.1967.J.Biol.Chem.242:2389-
2396.
126.Hoober,K.,P.Siekevitz,andG.E .Palade .1969.J.Biol.Chem.244:2621-
2631.
127 .Schor,S.,P.Siekevitz,andG.E.Palade.1970.Proc.Nail.Acad.Sci.U.S.
A.66:174-179.
128.Back,K.W.,P.Siekevitz,andG.E.Palade.1971 .J.Biol.Chem.246:188-
195.
129.Steere,R.L.1957.J.Biophys.Biochem.Cytol.3:45-60.
130.Moor,H.,K.Muhlethaler,H .Waldner,andA.Frey-Wysshng.1961.J.Cell
Biol.10:1-15.
131.Branton,D.1966.Proc.Nat.Acad.Sci.U.S.A.55:1048-1056.
132 .Branton,D.,andR.B.Park.1967.J.Ultrastruct.Res.19 :283-303.
133.Branton,D.1971.Philos.Trans:RSoc.Land.B.Biol.Sci.261:193-138.
134.daSilva,P.,andBranton,D .1970.J.Cell Biol.45:598-605.
135.Branton,D.1967.Exp.CellRes.45:703-707.
136.Branton,D.,andD.W.Deamer.1972.Protoplasmatologia.2(Sect.1):1-70.
137.Deamer,D.,andD.Branton .1967.Science(Wash.D.C).158:655-656.
138.Deamer,D.,R.Leonard,A.Tardieu,andD.Branton.1970 .Biochem.
Biophys.Acta.219:47-60.
139 .Mateu,L.,V.Luzzati,Y.London,R.M.Gould,andF.G.A.Vasseberg.
1973.J.Mal.Biol.75:697-709.
140.Gulik-Krzywieki,T.1975.Biochem.Biophys.Acta.415:1-28.
141.Clark,A.W.,andD.Branton.1968.Fracturefacesinfrozenoutersegments
fromtheguineapigretina.ZZellforsch.Mikrosk.Anal.91:586-603.
142.Leeson,T .S.1970.Ratretinalrods:freeze-fracture replicasofouter
segments.Can.J.Ophthalmol.5:91-107.
143.Leeson,T.S.1971.J.Anat.108:147-157.
144.Corless,J.M.,W.H.CobbsIII,M.J.Costello,andJ.D.Robertson.1976.
Exp.Eye Res.23:295-324.
145.Corless,J.M.,andM.J.Costello,1980.Biophys.J.25:289A.
146.Singer,S.J.,andG.L.Nicholson .1972.Science(Wash .D.C.).175:720-
731.
147.Frye,L,D.,andM.Edidin.1970.J.CellSci.7:319-335.
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
148.Evans,E.A.1973.Biophys.J.13:941.
149.LaCelle,P.1975.Biophys.J.15:210a.
150.Singer,S.J.1974.Int.Congr.ElectronMicrosc.Proc.8th.Canberra.186.
151.Nicholson,G .L.,V .T.Marchesi,andS.J.Singer.1971 .J.CellBiol.51:
265-272.
152.Zwaal,R.A.,andB.Roelofsen.1976.InBiochemicalAnalysisofMem-
branes.A.H.Maddy,editor.John WileyandSons,Inc.,NewYork.352-
377
.
153.Ottolenghi,A.1973.Lipids.8:415-425.
154.Adamich,M.,andE.A.Dennis.1979.InNormalandAbnormalRedCell
Membranes .S.E.Lux,V.T.Marchesi,andC.F.Fox,editors.AlanR.
Liss,Inc.,NewYork.515-521.
155.Yeager,M.,B.Schoenborn,D.Engelman,P.Moore,andL.Stryer.1980.
J.Mol.Biol.137:315-348.
156.Weinstein,R.S.,J,K.Khodadad,andT.L.Steck.1978.J.Supramol.
Struct.8:325-335.
157.Unwin,P.N.T.,andR.Henderson.1975.J.MotBiol.94:425-440.
158.Henderson,R.,andP.N.T.Unwin.1975.Nature (Lond).257:28-32.
159.Henderson,R.1975.J.Mol.Biol.93:123-138.
160.Jakoi,E.R.,G.Zampighi,andJ.D.Robertson.1976.InMembranesand
Disease.L.Bolis,J,F.Hoffman,andA.Leaf,editors.RavenPress,New
York.263-272.
161.Jakoi,E .R.,G.Zampighi,andJ.D.Robertson.1975.J.CellBiol.70:97-
111.
162.Robertson,J.D.,S.Knutton,A.R.Limbrick,E.R.Jakoi,andG.Zampighi.
1976.J.CellBiol.70:112-122.
163.Jakoi,E.R.,andR.B.Marchase.1979.J.CellBiol.80:642-650.
164.Fairbanks,G.,T.L.Steck,andD.F.A.Wallach.1971.Biochemistry.10:
2606-2617.
165.Steck,T,L.1978.InMembraneTransportProcesses,Vol .2.D.C.Tosteson,
Y.A.Ouchinnikov,andRamonLatorre,editors.RavenPress,NewYork.
39-51.
166.Steck,T.L.1972.J.MotBiol.66:295-304.
167.Steck,T.L.,andJ.Yu.1972.J.Supramol.Struct. 1:220-232.
168.Steck,T.L.,G.Fairbanks,andD.F.H.Wallach.1971.Biochemistry10:
2617-2624.
169.Steck,T.L.1978.J.Supramol.Struct.8:311-324.
170.Marchesi,V .T.,T.W.Tillack,R.L.Jackson,J.P.Segrest,andR .E.Scott.
1972.Proc.Nall.Acad.Sci.U.S.A.69:1445.
171.Bretscher,M.S.1972.J.MotBiol.71:523-528.
172.Bretscher,M.S.1971.Nat.NewBiol.231:229-232.
173.Bretscher,M.S.1973.Science(Wash.D,C.).181:622-629.
174.Whitley,H .M.,andH.C.Berg.1974.J.MotBiol.87:541-561.
175.Steck,T.L.,G.Fairbanks,andD.H.H.Wallach,1971.Biochemistry10:
2617-2624.
176.Bender,W.W.,H.Garan, andH.C .Berg.1971.J.Mol.Biol.58:783-797.
177.Triplett,R.B.,andK.L .Caraway.1972.Biochemistry.11:2897-2903.
178.Marchesi,V.T.,andE.Steers.1968.Science(Wash.D.C.).159:203.
179.Hsu,C .J.,A.Lomey,Y.Eshdat,andV.T.Marchesi.1979.J.Supramol.
Struct.10:227-239.
180.Maddy,A.H.,andM.J.Dunn.1978.J.SupramolStruct.8:465-471.
181.Mazia,D.,andA.Ruby.1968.Proc.Nall.Acad.Sci.U.S.A.61:1005-
1012.
182.Clarke,M.1971.Biochem.Biophys.Res.Commun .45:1063-1070.
183.Tanner,M.J .A.,andW.R.Grey.1971.Biochem.J.125:1109-1117.
184.Caraway,K .L.,andB.C.Shin.1973.J.Biol.Chem.247:2102-2108.
185.Tilney,L.G.,andP.Detmers.1975 .J.CellBiol.66:508-520 .
186.Gahmberg,C.G.,G.Tauren,1.Virtanen,and J.Wartiovaara .1978.J.
SupramolStruct.8:337-347 .
187.Furthmayer,H.1978 .J.Supramol.Struct.9:79-95.
188.Winzler,R.J.1969.InRedCellMembrane.G.A.JamiesonandT.J.
Greenwalt,editors.J.B.LippincottCo.,Philadelphia.157-171.
189.Morawiecki,A.1964.Biochim.Biophys.Acta.83:339-347 .
190.Marchesi,V.T.,andE.P.Andrews.1971.Science(Wash.D.C.).174:1247-
1248.
191.Cotmore,S.F.,H.Furthmayr,andV.T.Marchesi.1977.J.Mol.Biol.113:
539-553.
192.Tomrita,M.,andV.T.Marchesi.1976.Proc.Nall.Acad.Sci.U.S.A.72:
2964-2968.
193.Lin,S-C.,andJ.Palek,1979.J.SupramolStruct.10:97-109.
194.Ruoho,J.,andJ.Kyte,1974.Proc.Nall.Acad.Sci.U.S.A.71:2352-2356.
195.Kyte,J.1975.J.Biol.Chem.250:7443-7449.
196.Kant,J.D.,andJ.L.Steck.1972.Nature(Lond) .240:24-27.
197.Costello,M.J.,andT.Gulik-Krzywicki.1976.Biochim.Biophys.Acta.455:
412-432.
198.Hong,K.,andW.L.Hubbel.1972.Proc.Nall.Acad.Sci.U.S.A.69:2617-
2621.
199.Deamer,D.W.,andN.Yamanaka .1975.Biophys.J.15:110-111.
200.Verkleij,A.J.,R.F.A.Zwaal,B.Roelofsen,P.Comfurius,D.Kastellijn,
andL.L.M.vanDeenen.1973.Biochim.Biophys.Acta.323:178-193.
201.Kirk,R.G.,andD.C.Tosteson .1973.J.Membr.Biol.12:273-285.
202.daSilva,P.,S.T.Douglas,andD.Branton.1971.Nature(Lond)232:194-
196.
203.Tillack,T.W.,R.E.Scott,andV.T.Marchesi.1972.J.Exp .Med.135:
1209-1227.
204.daSilva,P.1972,JCellBiot53:777-787.
205.Elgsaeger,A.,andD .Branton.1974.J.CellBiol.63:1018-1030.
206.Asano,A.,andK.Aekiguchi.1978.J.Supramol.Struct.9:441-452.
207.Fujimoto,T.,andOgawa,K.1980.ActaHistochem.Cytochem.13 :72-89.
208.Pinder,J.C.,E.Ungwickell,D.Bray,andW.B.Gratzer .1978.J.Supramol.
Struct.8:439-445.
209.Baechi,T.,K.Whiting,M.J.A .Tanner,M.N.Metaxas,andD.J.Anstee.
1977.Biochem.Biophys.Acta.64:635-639 .
210.Bretscher,M.S.1973 .Science(Wash.D.C.).181:622-629 .
211.Gordesky, S.E.,andG.V.Marinetti.1973 .Biochem.Biophys.Res.Commun.
50:1027-1031.
212.Rothman,J.E.,andE.P.Kennedy.1977 .J.Mol.Biol.110:603-618.
213.Marinetti,E.V.,andR.C.Crain.1978.J.Supramol.Struct .8:191-213.
214 .VanDeenen,L.L.M.1979.InNormalandAbnormalRedCellMembranes.
S.E .Lux,V.T.Marchesi,andC.F.Fox,editors.AlanR.Liss,Inc.,New
York.451-456.
215.Verkleij,A.J.,R.F.A.Zwaal,B.Roelofsen,P.Comfurious,P .Kastelijn,
andL .L.M.VanDeenen,1973 .Biochim.Biophys.Acta.323:178-193.
216.Bloj,B.,andD.B.Zilversmit.1976 .Biochemistry.15:1277-1283.
217.Rouser,G.,G.J.Nelson,S.Flischer,andG.Simon.1968.In Biological
Membranes,PhysicalFact,andFunction.D.Chapman,editor .Academic
Press,Inc.,LondonandNewYork .5-69.
218.Kornberg,R.D.,andH.M.McConnell.1971.Biochemistry.10:1111-1120.
219.Kornberg,R.D.,andH.M.McConnell.1971.Proc.Nall.Acad.Sci.U.S.
A.68:2564-2568.
220.Fisher,K,A.1975.Science(Wash.D .C.)190:983-985.
221.Fisher,K.A.1976.Proc.Nall.AcadSci.U.S.A .73:173-177.
222.Dervichian,D.G.1954.InTheChemistry ofFatsandOtherLipids,Vol.
2 .R.T.Holman,W.O.Lundberg,andT.Malkin,editors .AcademicPress,
Inc.,NewYork.193.
223.Rand,R.P.,andV.Luzzati.1968.Biophys.J.8:125-137.
224.VanDeenen,L.L .M.,U .M.T .Houtsmuller,G.H.DeHaas,andE.
Mulder.1962.J.Pharm.Pharmacol14:429-444.
225.Shimshick,E .J.,andH.M.McConnell .1973.Biochemistry .12:2351-2360.
226.Ranck,J .L.,L.Mateu,D.M.Sadler,A.Tardieu,T .Gulik-Krzywicki,and
V.Luzzati.1974.J.MotBiol.85:249-277.
227.Klausner,R.D.,A.M.Kleinteld,R.L.Hoover,andM.J.Karnofsky.1980.
JBiol.Chem.255:1286-1295 .
228.Robertson,J.D.,andJ .A.Vergara.1980.Anal.Rec.196:159A.
229.Vergara,J.A.,A.Bakardjieva,E.Helmreich,and J.D.Robertson.1980.
Anal.Rec.196:195A-196A.
230.Robertson,J.D.,andJ.A.Vergara.1980.Electron.Microsc.Soc .Am.Proc.
38thAnnu.Meet.Inpress.
231.Robertson,J.D.,andJ.A.Vergara.1980.J.CellBiol.86:514-528.
232.Poo,Mu-Ming,Wen-jenPoo,andJ.W.Ham.1978 .J.CellBiol.76:483-
501.
233.Bourguignen .L.Y.W.,andS,J.Singer.1977.Proc.Nall.Acad.SciU.S.
A.74:5031-5035 .
234.Ryan,G.B.,J .Z.Borysenko,andM.J.Karnofsky.J.CellBiol.62:351-
365.
235.Schreiner,G.F.,andE.R.Unanue.1976.Adv .Immunol 24:38-165 .
236.Brown, J.,K.Fujiwara,J .D.Pollard,andE.R.Unanaw.1978.J.CellBiol.
79:409-418.
237.Brown, J.,K.Fujiwara,J.D.Pollard,andE.R.Unanaw.J.CellBiol.79:
419-426.
238.Stceckenius,W.,andW.H.Kunau.1968.J.CellBiot38:337-357.
239.Stoeckenius,W.,andR.H.Lozier.1974.J.Supramol.Struct.2:769-774.
240.Oesterhelt,D.,andW.Stceckenius,1971.Nat.NewBiol.233:149-152.
241.Oesterhelt,D.,andW.Stoeckenius.1973.Proc.Nall.Acad.Sci.U.S.A.70:
2853-2857.
242.Oesterhelt,D.,andW.Stoeckenius.1974.MethodsEnzymot3:667-678 .
243.Racket,E.,andW.Stoeckenius.1974.J.Biol.Chem.249:662-663.
244.Kates,M.,B.Palameta,C.N.Joo,D.J.Kushner,andN.E.Gibbons.1966.
Biochemistry.5:4092-4112.
245.Reynolds,J.A.,andH .Trayer .1971.J.Biol.Chem.246:7337-7342.
246.Fisher,K.A.,andW.Stoeckenius.1977.Science(Wash.D.C.).197:72-74.
247.Kuebler,O.,andH.Gross.1978.Int.Congr.ElectronMicrosc.Proc.9th.
Toronto.2:142-143.
248.Usukura,J.,E.Yamada,F.Tokunaga,andT.Yoshizawa.1980.J.Ultra-
struct.Res.70:204-219.
249.Robertson,J.D.,W.Schreil,andM.Reedy.1980.InElectronMicroscopy.
ElectronMicroscopySocietyofAmerica.W.Bailey,editor,Claitor'sPub-
lishingCo.,BocaRaton,Fla.792-793.
250.Engelman,D.M.,R.Henderson,A .D.McLachlan,andB.A.Wallace .
1980.Proc.Nall.Acad.Sci.U.S.A.77:2023-2037.
251.Ovchinikov,Y.A.,N.G .Abdulaev,M.Feigina,A.V.Kiselev,andN.A .
Oobanov.1979.FEBS(Fed.Eur.Biochem.Soc.)Lett.100:219-224,
252.Gerber.G.E.,R. J.Anderegg,W.C.Herlihy,C.P.Gray,K.Biesmann,
andH.G.Khorana.1979.Proc.Natl.Acad.Sci.U.S.A.76:227-231.
253.Walker,J.E.,A.F.Came,andH.W.Schmitt.1979.Nature (Loud.)278:
653-654.
254.Zaccai,G.,andD.M.Engleman .1979.J.Mol.Biol.132:181-191.
255.Engelman,D.M.1980.Fed.Proc.39:1964.
256.Bigelow,C.C.1967.J.Theor.Biol.16:187-211 .
257.Sjostrand,F.S.,andJ.Rhodin.1953.Exp.CellRes.4:426-456.
258.Sjostrand,F.S.,andV .Hanzon.1954.Exp.CellRes.2:393-414.
259.Sjostrand,F.S.,andE.Andersson.1954.Experientia(Basel).10:369-370.
ROBERTSON
MembraneStructure
2035
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
149.LaCelle,P.1975.Biophys.J.15:210a.
150.Singer,S.J.1974.Int.Congr.ElectronMicrosc.Proc.8th.Canberra.186.
151.Nicholson,G .L.,V .T.Marchesi,andS.J.Singer.1971 .J.CellBiol.51:
265-272.
152.Zwaal,R.A.,andB.Roelofsen.1976.InBiochemicalAnalysisofMem-
branes.A.H.Maddy,editor.John WileyandSons,Inc.,NewYork.352-
377
.
153.Ottolenghi,A.1973.Lipids.8:415-425.
154.Adamich,M.,andE.A.Dennis.1979.InNormalandAbnormalRedCell
Membranes .S.E.Lux,V.T.Marchesi,andC.F.Fox,editors.AlanR.
Liss,Inc.,NewYork.515-521.
155.Yeager,M.,B.Schoenborn,D.Engelman,P.Moore,andL.Stryer.1980.
J.Mol.Biol.137:315-348.
156.Weinstein,R.S.,J,K.Khodadad,andT.L.Steck.1978.J.Supramol.
Struct.8:325-335.
157.Unwin,P.N.T.,andR.Henderson.1975.J.MotBiol.94:425-440.
158.Henderson,R.,andP.N.T.Unwin.1975.Nature (Lond).257:28-32.
159.Henderson,R.1975.J.Mol.Biol.93:123-138.
160.Jakoi,E.R.,G.Zampighi,andJ.D.Robertson.1976.InMembranesand
Disease.L.Bolis,J,F.Hoffman,andA.Leaf,editors.RavenPress,New
York.263-272.
161.Jakoi,E .R.,G.Zampighi,andJ.D.Robertson.1975.J.CellBiol.70:97-
111.
162.Robertson,J.D.,S.Knutton,A.R.Limbrick,E.R.Jakoi,andG.Zampighi.
1976.J.CellBiol.70:112-122.
163.Jakoi,E.R.,andR.B.Marchase.1979.J.CellBiol.80:642-650.
164.Fairbanks,G.,T.L.Steck,andD.F.A.Wallach.1971.Biochemistry.10:
2606-2617.
165.Steck,T,L.1978.InMembraneTransportProcesses,Vol .2.D.C.Tosteson,
Y.A.Ouchinnikov,andRamonLatorre,editors.RavenPress,NewYork.
39-51.
166.Steck,T.L.1972.J.MotBiol.66:295-304.
167.Steck,T.L.,andJ.Yu.1972.J.Supramol.Struct. 1:220-232.
168.Steck,T.L.,G.Fairbanks,andD.F.H.Wallach.1971.Biochemistry10:
2617-2624.
169.Steck,T.L.1978.J.Supramol.Struct.8:311-324.
170.Marchesi,V .T.,T.W.Tillack,R.L.Jackson,J.P.Segrest,andR .E.Scott.
1972.Proc.Nall.Acad.Sci.U.S.A.69:1445.
171.Bretscher,M.S.1972.J.MotBiol.71:523-528.
172.Bretscher,M.S.1971.Nat.NewBiol.231:229-232.
173.Bretscher,M.S.1973.Science(Wash.D,C.).181:622-629.
174.Whitley,H .M.,andH.C.Berg.1974.J.MotBiol.87:541-561.
175.Steck,T.L.,G.Fairbanks,andD.H.H.Wallach,1971.Biochemistry10:
2617-2624.
176.Bender,W.W.,H.Garan, andH.C .Berg.1971.J.Mol.Biol.58:783-797.
177.Triplett,R.B.,andK.L .Caraway.1972.Biochemistry.11:2897-2903.
178.Marchesi,V.T.,andE.Steers.1968.Science(Wash.D.C.).159:203.
179.Hsu,C .J.,A.Lomey,Y.Eshdat,andV.T.Marchesi.1979.J.Supramol.
Struct.10:227-239.
180.Maddy,A.H.,andM.J.Dunn.1978.J.SupramolStruct.8:465-471.
181.Mazia,D.,andA.Ruby.1968.Proc.Nall.Acad.Sci.U.S.A.61:1005-
1012.
182.Clarke,M.1971.Biochem.Biophys.Res.Commun .45:1063-1070.
183.Tanner,M.J .A.,andW.R.Grey.1971.Biochem.J.125:1109-1117.
184.Caraway,K .L.,andB.C.Shin.1973.J.Biol.Chem.247:2102-2108.
185.Tilney,L.G.,andP.Detmers.1975 .J.CellBiol.66:508-520 .
186.Gahmberg,C.G.,G.Tauren,1.Virtanen,and J.Wartiovaara .1978.J.
SupramolStruct.8:337-347 .
187.Furthmayer,H.1978 .J.Supramol.Struct.9:79-95.
188.Winzler,R.J.1969.InRedCellMembrane.G.A.JamiesonandT.J.
Greenwalt,editors.J.B.LippincottCo.,Philadelphia.157-171.
189.Morawiecki,A.1964.Biochim.Biophys.Acta.83:339-347 .
190.Marchesi,V.T.,andE.P.Andrews.1971.Science(Wash.D.C.).174:1247-
1248.
191.Cotmore,S.F.,H.Furthmayr,andV.T.Marchesi.1977.J.Mol.Biol.113:
539-553.
192.Tomrita,M.,andV.T.Marchesi.1976.Proc.Nall.Acad.Sci.U.S.A.72:
2964-2968.
193.Lin,S-C.,andJ.Palek,1979.J.SupramolStruct.10:97-109.
194.Ruoho,J.,andJ.Kyte,1974.Proc.Nall.Acad.Sci.U.S.A.71:2352-2356.
195.Kyte,J.1975.J.Biol.Chem.250:7443-7449.
196.Kant,J.D.,andJ.L.Steck.1972.Nature(Lond) .240:24-27.
197.Costello,M.J.,andT.Gulik-Krzywicki.1976.Biochim.Biophys.Acta.455:
412-432.
198.Hong,K.,andW.L.Hubbel.1972.Proc.Nall.Acad.Sci.U.S.A.69:2617-
2621.
199.Deamer,D.W.,andN.Yamanaka .1975.Biophys.J.15:110-111.
200.Verkleij,A.J.,R.F.A.Zwaal,B.Roelofsen,P.Comfurius,D.Kastellijn,
andL.L.M.vanDeenen.1973.Biochim.Biophys.Acta.323:178-193.
201.Kirk,R.G.,andD.C.Tosteson .1973.J.Membr.Biol.12:273-285.
202.daSilva,P.,S.T.Douglas,andD.Branton.1971.Nature(Lond)232:194-
196.
203.Tillack,T.W.,R.E.Scott,andV.T.Marchesi.1972.J.Exp .Med.135:
1209-1227.
204.daSilva,P.1972,JCellBiot53:777-787.
205.Elgsaeger,A.,andD .Branton.1974.J.CellBiol.63:1018-1030.
206.Asano,A.,andK.Aekiguchi.1978.J.Supramol.Struct.9:441-452.
207.Fujimoto,T.,andOgawa,K.1980.ActaHistochem.Cytochem.13 :72-89.
208.Pinder,J.C.,E.Ungwickell,D.Bray,andW.B.Gratzer .1978.J.Supramol.
Struct.8:439-445.
209.Baechi,T.,K.Whiting,M.J.A .Tanner,M.N.Metaxas,andD.J.Anstee.
1977.Biochem.Biophys.Acta.64:635-639 .
210.Bretscher,M.S.1973 .Science(Wash.D.C.).181:622-629 .
211.Gordesky, S.E.,andG.V.Marinetti.1973 .Biochem.Biophys.Res.Commun.
50:1027-1031.
212.Rothman,J.E.,andE.P.Kennedy.1977 .J.Mol.Biol.110:603-618.
213.Marinetti,E.V.,andR.C.Crain.1978.J.Supramol.Struct .8:191-213.
214 .VanDeenen,L.L.M.1979.InNormalandAbnormalRedCellMembranes.
S.E .Lux,V.T.Marchesi,andC.F.Fox,editors.AlanR.Liss,Inc.,New
York.451-456.
215.Verkleij,A.J.,R.F.A.Zwaal,B.Roelofsen,P.Comfurious,P .Kastelijn,
andL .L.M.VanDeenen,1973 .Biochim.Biophys.Acta.323:178-193.
216.Bloj,B.,andD.B.Zilversmit.1976 .Biochemistry.15:1277-1283.
217.Rouser,G.,G.J.Nelson,S.Flischer,andG.Simon.1968.In Biological
Membranes,PhysicalFact,andFunction.D.Chapman,editor .Academic
Press,Inc.,LondonandNewYork .5-69.
218.Kornberg,R.D.,andH.M.McConnell.1971.Biochemistry.10:1111-1120.
219.Kornberg,R.D.,andH.M.McConnell.1971.Proc.Nall.Acad.Sci.U.S.
A.68:2564-2568.
220.Fisher,K,A.1975.Science(Wash.D .C.)190:983-985.
221.Fisher,K.A.1976.Proc.Nall.AcadSci.U.S.A .73:173-177.
222.Dervichian,D.G.1954.InTheChemistry ofFatsandOtherLipids,Vol.
2 .R.T.Holman,W.O.Lundberg,andT.Malkin,editors .AcademicPress,
Inc.,NewYork.193.
223.Rand,R.P.,andV.Luzzati.1968.Biophys.J.8:125-137.
224.VanDeenen,L.L .M.,U .M.T .Houtsmuller,G.H.DeHaas,andE.
Mulder.1962.J.Pharm.Pharmacol14:429-444.
225.Shimshick,E .J.,andH.M.McConnell .1973.Biochemistry .12:2351-2360.
226.Ranck,J .L.,L.Mateu,D.M.Sadler,A.Tardieu,T .Gulik-Krzywicki,and
V.Luzzati.1974.J.MotBiol.85:249-277.
227.Klausner,R.D.,A.M.Kleinteld,R.L.Hoover,andM.J.Karnofsky.1980.
JBiol.Chem.255:1286-1295 .
228.Robertson,J.D.,andJ .A.Vergara.1980.Anal.Rec.196:159A.
229.Vergara,J.A.,A.Bakardjieva,E.Helmreich,and J.D.Robertson.1980.
Anal.Rec.196:195A-196A.
230.Robertson,J.D.,andJ.A.Vergara.1980.Electron.Microsc.Soc .Am.Proc.
38thAnnu.Meet.Inpress.
231.Robertson,J.D.,andJ.A.Vergara.1980.J.CellBiol.86:514-528.
232.Poo,Mu-Ming,Wen-jenPoo,andJ.W.Ham.1978 .J.CellBiol.76:483-
501.
233.Bourguignen .L.Y.W.,andS,J.Singer.1977.Proc.Nall.Acad.SciU.S.
A.74:5031-5035 .
234.Ryan,G.B.,J .Z.Borysenko,andM.J.Karnofsky.J.CellBiol.62:351-
365.
235.Schreiner,G.F.,andE.R.Unanue.1976.Adv .Immunol 24:38-165 .
236.Brown, J.,K.Fujiwara,J .D.Pollard,andE.R.Unanaw.1978.J.CellBiol.
79:409-418.
237.Brown, J.,K.Fujiwara,J.D.Pollard,andE.R.Unanaw.J.CellBiol.79:
419-426.
238.Stceckenius,W.,andW.H.Kunau.1968.J.CellBiot38:337-357.
239.Stoeckenius,W.,andR.H.Lozier.1974.J.Supramol.Struct.2:769-774.
240.Oesterhelt,D.,andW.Stceckenius,1971.Nat.NewBiol.233:149-152.
241.Oesterhelt,D.,andW.Stoeckenius.1973.Proc.Nall.Acad.Sci.U.S.A.70:
2853-2857.
242.Oesterhelt,D.,andW.Stoeckenius.1974.MethodsEnzymot3:667-678 .
243.Racket,E.,andW.Stoeckenius.1974.J.Biol.Chem.249:662-663.
244.Kates,M.,B.Palameta,C.N.Joo,D.J.Kushner,andN.E.Gibbons.1966.
Biochemistry.5:4092-4112.
245.Reynolds,J.A.,andH .Trayer .1971.J.Biol.Chem.246:7337-7342.
246.Fisher,K.A.,andW.Stoeckenius.1977.Science(Wash.D.C.).197:72-74.
247.Kuebler,O.,andH.Gross.1978.Int.Congr.ElectronMicrosc.Proc.9th.
Toronto.2:142-143.
248.Usukura,J.,E.Yamada,F.Tokunaga,andT.Yoshizawa.1980.J.Ultra-
struct.Res.70:204-219.
249.Robertson,J.D.,W.Schreil,andM.Reedy.1980.InElectronMicroscopy.
ElectronMicroscopySocietyofAmerica.W.Bailey,editor,Claitor'sPub-
lishingCo.,BocaRaton,Fla.792-793.
250.Engelman,D.M.,R.Henderson,A .D.McLachlan,andB.A.Wallace .
1980.Proc.Nall.Acad.Sci.U.S.A.77:2023-2037.
251.Ovchinikov,Y.A.,N.G .Abdulaev,M.Feigina,A.V.Kiselev,andN.A .
Oobanov.1979.FEBS(Fed.Eur.Biochem.Soc.)Lett.100:219-224,
252.Gerber.G.E.,R. J.Anderegg,W.C.Herlihy,C.P.Gray,K.Biesmann,
andH.G.Khorana.1979.Proc.Natl.Acad.Sci.U.S.A.76:227-231.
253.Walker,J.E.,A.F.Came,andH.W.Schmitt.1979.Nature (Loud.)278:
653-654.
254.Zaccai,G.,andD.M.Engleman .1979.J.Mol.Biol.132:181-191.
255.Engelman,D.M.1980.Fed.Proc.39:1964.
256.Bigelow,C.C.1967.J.Theor.Biol.16:187-211 .
257.Sjostrand,F.S.,andJ.Rhodin.1953.Exp.CellRes.4:426-456.
258.Sjostrand,F.S.,andV .Hanzon.1954.Exp.CellRes.2:393-414.
259.Sjostrand,F.S.,andE.Andersson.1954.Experientia(Basel).10:369-370.
ROBERTSON
MembraneStructure
2035
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
260.Sjostrand,F.S.1953.Nature(Land.).171 :30.
261.Sjostrand,F.S.,E.Andersson-Cedergren,andM.M.Dewey.1958.J.
Ultrastruct.Res.1:271-287.
262.Robertson,J.D.1959.Z.Zel!(orsch.Mikrosk .Anat.50:553-560.
263.Robertson,J.D.1953.Proc.Soc.Exp.Biol.Med.82:219-223.
264.Robertson,J.D.1961.Ann.N.Y.Acad.Sci.94:339-389.
265.Furshpan,E.J.,andD.D.Potter.1959.J.Physiol(Land.).145:289-325.
266.Karrer,H.E.,andJ.Cox.1960.J.Biophys.Biochem.Cytol 8:135-150 .
267.Furshpan,E.J.,andT.Furakawa.1962.J.Neurophysiol25:732-771 .
268.Furshpan,E.J.1964.Science(Wash.D.C.).144:878-880 .
269.Bartelmez,G.W.1915.J.Comp.Neural.25:87-128.
270.Robertson,J.D.,T.S.Bodenheimer,andD.E.Stage.InProceedingsofthe
AmericanSocietyforCellBiology2ndAnnualMeeting,SanFrancisco.
194.
271.Robertson,J.D.,T.S.Bodenheimer,andD.E.Stage.1963.J.CellBiol.19 :
159-200.
272.Robertson,J.D.1963.J.CellBiol.19 :201-221.
273.Goodenough,D.A.,andJ.P.Revel.1970.J.CellBiol.45:272-290.
274.Dewey,M.M.,andL.Barr.1962 .Science(Wash.D.C).137:670-672.
275.Dewey,M.M.,andL.Barr.1964 .J.CellBiol.23:553-585.
276.Benedetti,E.L.,andP.Emmelot.1965.J.CellBiol.26:299-304.
277.Benedetti,E .L.,andP.Emmelot.1968.J.CellBiol.38:15-24.
278.Farquhar,M.G.,andG.E .Palade .1963.J.CellBiol.17:375-412.
279.Farquhar,M.G.,andG.E.Palade.1964.Proc.Nat.Acad.Sci.U.S.A.51:
569-577.
280.Farquhar,M.G.,andG.E .Palade.1965.J.CellBiol.26:263-291.
281.Knutton,S.,A.R.Limbrick,and J.D.Robertson.1978.CellTissueRes.
191:449-462.
282.Lettvin,J.Y.,W.F.Pickard,J.W.Moore, andM.Takata.1964 .Quarterly
ProgressReport,ResearchLaboratoryofMIT.75:159.
283.Doggenweiler,C.F.,andS.Freak.1965.Proc.Nall.Acad.Sci.U.S.A.53:
425-430.
284.Revel,J.P.,andM.J.Karnowski.1967.J.CellBiol.33:C7-C12.
285.Simionescu,M.,N.Simionescu,andG.E.Palade.1975.J.CellBiol.67 :
863-885.
286.Kreutziger,G.O.1968.ElectronMicrosc.Soc.Am.Proc.16thAnnu.Meet.
234-235.
287.Bullivant,S.1969.ElectronMicrosc.Soc.Am.Proc.17thAnnu.Meet.206-
207
.
288.McNutt,N.S.,andR.S.Weinstein.1969.ElectronMicrosc.Soc.Am.Proc.
17thAnnu.Meet.330.
289.Chaleroft,J.P.,and S.Bullivant.1970.J.CellBiol.47:49-60.
290.Steere,R .L.,andJ.R.Sommer.1972.J.Microsc .(Paris).15:205-218.
291.McNutt,N.S.,andR.S.Weinstein.1970.J.CellBiol.47:666-688.
292.Peracchia,C.1973.J.CellBiol.57:54--65.
293.Peracchia,C.1973.J.CellBiol.57:66-76.
294.Peracchia,C.,andA.S .Dulhunty.1976.J.CellBiol.70:419-439.
295.Peracchia,C.1977.J.CellBiol.72:628-641.
296.Lowenstein,W.R.1966.Ann.N.Y.Acad.Sci.137:441-472 .
204S
THEJOURNALOF CELLBIOLOGY"VOLUME91,1981
297.Potter,D.D.,E.J.Furshpan,andE.S.Lennox.1966.Proc.Nall.AcadSci.
U.S.A.55:328-336.
298.Pappas,G.D.,andM.V.L.Bennett.1966.Ann.N.YAcad.Sci.137:495-
508.
299.Sheridan,J .D.1971.Dev.Biol.26:627-636.
300.Lowenstein,W.1979 .InProceedingsoftheInternationalConferenceon
theBiologyoftheMembrane,CranssurSierre,D.C.Tosteson,L.Bolis,K.
Bloch,editors.Inpress.
301 .Warner,A.1979.InProceedingsoftheInternationalConferenceonthe
BiologyoftheMembrane,CranssurSierre,D.C.Tosteson,L.Bolis,K.
Bloch,editors.Inpress.
302.Goodenough,D.A.,andW.Stoeckenius.1972.J.CellBiol.54:646-656.
303.Evans,W.H.,andJ.W.Gurd.1972.Biochem.J.128:691-700.
304.Goodenough,D.A.1974.J.CellBiol.61:557-563.
305.Gilula,N.B.1974.J.CellBiol.63:111a .
306.Dunia,I.,C.Sen Oshosh,E.L.Benedetti,A.Zeers,andH.Bloemendal.
1974 .FEBS(Fed.Eur.Biochem.Soc.)Lett.45:139-144.
307.Goodenough,D.A.1976.J.CellBiol.68:220-231.
308.Goodenough,D.A .1975.ColdSpringHarborSymp.Quant.Biol.40:37-
43.
309.Dugoid, J.R.,andJ-P.Revel.1975.ColdSpringHarborSymp.Quant.Biol.
40:45-47.
310.Benedetti,E.L.,I .Dunia,C.J.Bentzel,A.J.M.VeLmorken,M.Kibbelaar,
andH.Bloemendal .1976.Biochem.Biophys.Acta.457:353-384 .
311.Culvenar,J.G.,andW.H.Evans.1977.Biochem.J.168:475-481.
312.Zampighi,G.A.,andJ.D .Robertson.1977.Biophys.J.13:71A .
313.Zampighi,G.A.1978.Ph.D.ThesisDissertation.DukeUniversityLibrary,
Durham,N.C.
314.Ehrhart,J.C.,andJ.Chauveau.1977 .FEBS(Fed.Eur.Biochem.Soc.)
Lett,78:295-299.
315.Gilula,N.B.1978.J.CellBiol.79(No.2,Part2):223a.
316.Hertzberg,E .L.,andN.B.Gilula.1979.J.Biol.Chem.254:2138-2147.
317.Henderson,D.,H .Eibl,andK .Weber.1979.J.MolBiol.132:193-218.
318.Barrantes,J.J.1975.Biochem.Biophys.Res.Commun.62:407-414.
319.Finbow,M.,S.B.Yancey,R.Johnson,andJ-P.Revel.1980 .Proc.Nad.
Acad.Sci.U.S.A.77:970-974.
320.Yee,A.G.,andJ-P.Revel.1978.J.CellBiol.78:554-564.
321.Yancey, S.B.,D.Easter,andJ-P.Revel.1979.J.Ultrastruct.Res.67:229-
242.
322.Caspar,D.L.D.,D.A.Goodenough,L.Makowski,andW.C.Phillips.
1977.J.CellBiol.74:605-628.
323.Makowski,L.,D.L.D .Caspar,W.C.Phillips,D.A.Phillips,andD.A.
Goodenough.1977.J.CellBiol.74:629-645.
324 .Goodenough,D.A.1976.J.CellBiel.71:334-335 .
325.Zampighi,G.,and J.D.Robertson .1973.J.CellBiol.56:92-105.
326.Brenner,S.,andR.W.Home.1959.Biochem .Biophys.Acta .34:103-110.
327 .Zampighi,G.,J.Corless,andRobertson,J.D.1980 .J.CellBiol.86:190.
328.Zampighi,G.,andP.N.T.Unwin.1979.J.MolBiol.135:541-564.
329.Unwin,P.N.T.,andG.Zampighi.1980.Nature(Loud.).283:545-549.
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
261.Sjostrand,F.S.,E.Andersson-Cedergren,andM.M.Dewey.1958.J.
Ultrastruct.Res.1:271-287.
262.Robertson,J.D.1959.Z.Zel!(orsch.Mikrosk .Anat.50:553-560.
263.Robertson,J.D.1953.Proc.Soc.Exp.Biol.Med.82:219-223.
264.Robertson,J.D.1961.Ann.N.Y.Acad.Sci.94:339-389.
265.Furshpan,E.J.,andD.D.Potter.1959.J.Physiol(Land.).145:289-325.
266.Karrer,H.E.,andJ.Cox.1960.J.Biophys.Biochem.Cytol 8:135-150 .
267.Furshpan,E.J.,andT.Furakawa.1962.J.Neurophysiol25:732-771 .
268.Furshpan,E.J.1964.Science(Wash.D.C.).144:878-880 .
269.Bartelmez,G.W.1915.J.Comp.Neural.25:87-128.
270.Robertson,J.D.,T.S.Bodenheimer,andD.E.Stage.InProceedingsofthe
AmericanSocietyforCellBiology2ndAnnualMeeting,SanFrancisco.
194.
271.Robertson,J.D.,T.S.Bodenheimer,andD.E.Stage.1963.J.CellBiol.19 :
159-200.
272.Robertson,J.D.1963.J.CellBiol.19 :201-221.
273.Goodenough,D.A.,andJ.P.Revel.1970.J.CellBiol.45:272-290.
274.Dewey,M.M.,andL.Barr.1962 .Science(Wash.D.C).137:670-672.
275.Dewey,M.M.,andL.Barr.1964 .J.CellBiol.23:553-585.
276.Benedetti,E.L.,andP.Emmelot.1965.J.CellBiol.26:299-304.
277.Benedetti,E .L.,andP.Emmelot.1968.J.CellBiol.38:15-24.
278.Farquhar,M.G.,andG.E .Palade .1963.J.CellBiol.17:375-412.
279.Farquhar,M.G.,andG.E.Palade.1964.Proc.Nat.Acad.Sci.U.S.A.51:
569-577.
280.Farquhar,M.G.,andG.E .Palade.1965.J.CellBiol.26:263-291.
281.Knutton,S.,A.R.Limbrick,and J.D.Robertson.1978.CellTissueRes.
191:449-462.
282.Lettvin,J.Y.,W.F.Pickard,J.W.Moore, andM.Takata.1964 .Quarterly
ProgressReport,ResearchLaboratoryofMIT.75:159.
283.Doggenweiler,C.F.,andS.Freak.1965.Proc.Nall.Acad.Sci.U.S.A.53:
425-430.
284.Revel,J.P.,andM.J.Karnowski.1967.J.CellBiol.33:C7-C12.
285.Simionescu,M.,N.Simionescu,andG.E.Palade.1975.J.CellBiol.67 :
863-885.
286.Kreutziger,G.O.1968.ElectronMicrosc.Soc.Am.Proc.16thAnnu.Meet.
234-235.
287.Bullivant,S.1969.ElectronMicrosc.Soc.Am.Proc.17thAnnu.Meet.206-
207
.
288.McNutt,N.S.,andR.S.Weinstein.1969.ElectronMicrosc.Soc.Am.Proc.
17thAnnu.Meet.330.
289.Chaleroft,J.P.,and S.Bullivant.1970.J.CellBiol.47:49-60.
290.Steere,R .L.,andJ.R.Sommer.1972.J.Microsc .(Paris).15:205-218.
291.McNutt,N.S.,andR.S.Weinstein.1970.J.CellBiol.47:666-688.
292.Peracchia,C.1973.J.CellBiol.57:54--65.
293.Peracchia,C.1973.J.CellBiol.57:66-76.
294.Peracchia,C.,andA.S .Dulhunty.1976.J.CellBiol.70:419-439.
295.Peracchia,C.1977.J.CellBiol.72:628-641.
296.Lowenstein,W.R.1966.Ann.N.Y.Acad.Sci.137:441-472 .
204S
THEJOURNALOF CELLBIOLOGY"VOLUME91,1981
297.Potter,D.D.,E.J.Furshpan,andE.S.Lennox.1966.Proc.Nall.AcadSci.
U.S.A.55:328-336.
298.Pappas,G.D.,andM.V.L.Bennett.1966.Ann.N.YAcad.Sci.137:495-
508.
299.Sheridan,J .D.1971.Dev.Biol.26:627-636.
300.Lowenstein,W.1979 .InProceedingsoftheInternationalConferenceon
theBiologyoftheMembrane,CranssurSierre,D.C.Tosteson,L.Bolis,K.
Bloch,editors.Inpress.
301 .Warner,A.1979.InProceedingsoftheInternationalConferenceonthe
BiologyoftheMembrane,CranssurSierre,D.C.Tosteson,L.Bolis,K.
Bloch,editors.Inpress.
302.Goodenough,D.A.,andW.Stoeckenius.1972.J.CellBiol.54:646-656.
303.Evans,W.H.,andJ.W.Gurd.1972.Biochem.J.128:691-700.
304.Goodenough,D.A.1974.J.CellBiol.61:557-563.
305.Gilula,N.B.1974.J.CellBiol.63:111a .
306.Dunia,I.,C.Sen Oshosh,E.L.Benedetti,A.Zeers,andH.Bloemendal.
1974 .FEBS(Fed.Eur.Biochem.Soc.)Lett.45:139-144.
307.Goodenough,D.A.1976.J.CellBiol.68:220-231.
308.Goodenough,D.A .1975.ColdSpringHarborSymp.Quant.Biol.40:37-
43.
309.Dugoid, J.R.,andJ-P.Revel.1975.ColdSpringHarborSymp.Quant.Biol.
40:45-47.
310.Benedetti,E.L.,I .Dunia,C.J.Bentzel,A.J.M.VeLmorken,M.Kibbelaar,
andH.Bloemendal .1976.Biochem.Biophys.Acta.457:353-384 .
311.Culvenar,J.G.,andW.H.Evans.1977.Biochem.J.168:475-481.
312.Zampighi,G.A.,andJ.D .Robertson.1977.Biophys.J.13:71A .
313.Zampighi,G.A.1978.Ph.D.ThesisDissertation.DukeUniversityLibrary,
Durham,N.C.
314.Ehrhart,J.C.,andJ.Chauveau.1977 .FEBS(Fed.Eur.Biochem.Soc.)
Lett,78:295-299.
315.Gilula,N.B.1978.J.CellBiol.79(No.2,Part2):223a.
316.Hertzberg,E .L.,andN.B.Gilula.1979.J.Biol.Chem.254:2138-2147.
317.Henderson,D.,H .Eibl,andK .Weber.1979.J.MolBiol.132:193-218.
318.Barrantes,J.J.1975.Biochem.Biophys.Res.Commun.62:407-414.
319.Finbow,M.,S.B.Yancey,R.Johnson,andJ-P.Revel.1980 .Proc.Nad.
Acad.Sci.U.S.A.77:970-974.
320.Yee,A.G.,andJ-P.Revel.1978.J.CellBiol.78:554-564.
321.Yancey, S.B.,D.Easter,andJ-P.Revel.1979.J.Ultrastruct.Res.67:229-
242.
322.Caspar,D.L.D.,D.A.Goodenough,L.Makowski,andW.C.Phillips.
1977.J.CellBiol.74:605-628.
323.Makowski,L.,D.L.D .Caspar,W.C.Phillips,D.A.Phillips,andD.A.
Goodenough.1977.J.CellBiol.74:629-645.
324 .Goodenough,D.A.1976.J.CellBiel.71:334-335 .
325.Zampighi,G.,and J.D.Robertson .1973.J.CellBiol.56:92-105.
326.Brenner,S.,andR.W.Home.1959.Biochem .Biophys.Acta .34:103-110.
327 .Zampighi,G.,J.Corless,andRobertson,J.D.1980 .J.CellBiol.86:190.
328.Zampighi,G.,andP.N.T.Unwin.1979.J.MolBiol.135:541-564.
329.Unwin,P.N.T.,andG.Zampighi.1980.Nature(Loud.).283:545-549.
on September 2, 2014jcb.rupress.orgDownloaded from
Published February 22, 1981
Tags
Categories
TechnologyDownload
Download Slideshow Get the original presentation fileQuick Actions
Statistics
- Views 1,385
- Slides 16
- Favorites 3
- Age 3531 days
Related Slideshows
11
8-top-ai-courses-for-customer-support-representatives-in-2025.pptx
JeroenErne2
86 views
10
7-essential-ai-courses-for-call-center-supervisors-in-2025.pptx
JeroenErne2
80 views
13
25-essential-ai-courses-for-user-support-specialists-in-2025.pptx
JeroenErne2
76 views
11
8-essential-ai-courses-for-insurance-customer-service-representatives-in-2025.pptx
JeroenErne2
62 views
21
Know for Certain
DaveSinNM
38 views
17
PPT OPD LES 3ertt4t4tqqqe23e3e3rq2qq232.pptx
novasedanayoga46
41 views